Skip to main content

Full text of "The face of the earth (Das antlitz der erde)"

See other formats


Google 



This is a digital copy of a book that was preserved for generations on library shelves before it was carefully scanned by Google as part of a project 

to make the world's books discoverable online. 

It has survived long enough for the copyright to expire and the book to enter the public domain. A public domain book is one that was never subject 

to copyright or whose legal copyright term has expired. Whether a book is in the public domain may vary country to country. Public domain books 

are our gateways to the past, representing a wealth of history, culture and knowledge that's often difficult to discover. 

Marks, notations and other maiginalia present in the original volume will appear in this file - a reminder of this book's long journey from the 

publisher to a library and finally to you. 

Usage guidelines 

Google is proud to partner with libraries to digitize public domain materials and make them widely accessible. Public domain books belong to the 
public and we are merely their custodians. Nevertheless, this work is expensive, so in order to keep providing tliis resource, we liave taken steps to 
prevent abuse by commercial parties, including placing technical restrictions on automated querying. 
We also ask that you: 

+ Make non-commercial use of the files We designed Google Book Search for use by individuals, and we request that you use these files for 
personal, non-commercial purposes. 

+ Refrain fivm automated querying Do not send automated queries of any sort to Google's system: If you are conducting research on machine 
translation, optical character recognition or other areas where access to a large amount of text is helpful, please contact us. We encourage the 
use of public domain materials for these purposes and may be able to help. 

+ Maintain attributionTht GoogXt "watermark" you see on each file is essential for in forming people about this project and helping them find 
additional materials through Google Book Search. Please do not remove it. 

+ Keep it legal Whatever your use, remember that you are responsible for ensuring that what you are doing is legal. Do not assume that just 
because we believe a book is in the public domain for users in the United States, that the work is also in the public domain for users in other 
countries. Whether a book is still in copyright varies from country to country, and we can't offer guidance on whether any specific use of 
any specific book is allowed. Please do not assume that a book's appearance in Google Book Search means it can be used in any manner 
anywhere in the world. Copyright infringement liabili^ can be quite severe. 

About Google Book Search 

Google's mission is to organize the world's information and to make it universally accessible and useful. Google Book Search helps readers 
discover the world's books while helping authors and publishers reach new audiences. You can search through the full text of this book on the web 

at |http: //books .google .com/I 



S^ar?. ?9 0J 



Sairbatt aoUeic lUitati 



DANIEL TREADWELL FUND 



Rciiduiry [egic]' from Daniel TnadwclJ, Rumford 

PtofMiot ind Lecturer on the Application 

nf Sdence to the Uicfii] Arts 

i8j4-iS«. 



THE FACE OF THE EARTH 

(DAS ANTLITZ DER ERDE) 

By EDUARD SUESS 

PBOFB880B OF GB0L06T IN THE UHITXBSmr OF YIXNKA 
FO&EIGN MSMBEB OF THS BOTAL 80CIBTT OF LONDOH 



TRAKSLATED BT 

HERTHA B. C. SOLLAS 

PH.D. HUDELBBRO; OF NEWNHAM COLLEGE^ CAMBRIDGE 

XJNBEB THE DIBBCTION OF 

W. J. SOLLAS 

aC.D. (cANTAB.), LL.D. (dUBLIN), M.A. (oXON.), F.R.8. 

FKLLOW OF UMITBIumr COLLBOE, OXFORD 

PROFESSOR OF OBOLOOY IN THE UNIVERSITY OF OXFORD 



VOL. II 



iD 



OXFORD 



AT THE CLARENDON PRESS 

1906 < 




II 



lU 



CONTENTS 

PART III 

The Sea 

PAGE 

Chapter I. Conflict of Opinion regarding the Displacement of the Strand : 
Terminology and General Observations. Strabo. Dante. Celuus and 
LinnaeuB. Nordenankar. Flay&ir and Leoi)old von Buch. Groeihe. Lyelland 
Darwin as rappoiters of the theory of elevation. Bravais and Engine Robert. 
Chambers and Domeyko; renewed expreadons of donbt Theories based on 
gravitation. AdhtoarandhiBsacceasors. Summaries; Howorth. Terminology. 
Genuine case of dislocation in Mew Zealand. Plan of the following chapters . 1 

Chapter n. The Outlines of the Atlantic Ocean. The Canadian shield. The 
Baltic shield. Glint lines. The table-land of Spitsbergen. Greenland. The 
Caledonian monntains. The Armorican mountains. The Variscan mountains. 
The ^yntaxis of Central Europe. The Iberian Meseta. Survey of the pre- 
Permian mountains in Europe. The islands of Europe. Western Africa. The 
east of Central and South America. Survey of the outlines of the Atlantic . 30 

Chapter m. The Borders of the Pacific Ocean. New Zealand. Australia. 
New Caledonia. The sea of Banda. Borneo. Cochin China, Tongking. The 
Philippines. Formosa and the Liu-Eiu islands. Japan. The Euriles and 
KauEichatha. General survey of the island arcs. Eastern China. North-eastern 
Asia. The arc of the Aleutian islands. The west coast of America . . • 148 

Chapter IV. Comparison of the Atlantic and Pacific Outlines. The Atlantic 
structure. The Pacific structure. Their dissimilarity. Distribution of islands 
and volcanos. Overthmsting of the depressions. Progressive completion of the 
Hescsoic series towards the Pacific coast 201 

Chapter V. Palaeozoic Seas. Introduction : The abyssal region ; cycles ; thick- 
ness of the sediments. The North Atlantic continent. Upper limit of the 
Silurian. Universality of the middle Devonian transgression. The Carboniferous 
system. Paralic beds. Alternation of coal beds and marine sediments. Trans- 
gression of the Carboniferous limestone. The Permian system. Gondw&na 
land. General Survey 206 

Chapter VL Mesosoic Seas. The seas of the Trias. Positive movements in 
the Rhaetic epoch. Continuation of positive movements during the Jurassic 
ei>och. Negative phase in central Europe and commencement of the Cre- 
taceous epoch. Later transgressions and intermixture of the Cretaceous fiftunas. 
The Cenomanian transgression. General survey of the Hesozoic seas • . 256 

Chapter Vn. Tertiary Seas and recent Limestone Formations. Negative 
phase at the close of the Cretaceous epoch. The central Mediterranean of the 
Tertiaiy aera. The east coast of North America. The Tertiary region of 
Patagonia. Becent limestone formations. Summaiy 296 

Chapter Vm. The Strand*lines of Norway. From Tjoalma Vagge to the sea. 
From lake Tome& to the sea. Movement of the ice contraiy to the slope of the 
valleys. Origin of the glint lakes of Lapland. Ancient strand-lines of the Qords. 
Origin of the seter. Comparison with Greenland. Vestiges left by the retreat- 
ing ice 826 



I 



^toL ssi. ' ^y ^f-r^-^ 




5^^^?,(f^ C^l 




• I 






JUL 7 1905 






ljCajtgCoLK.CC 



J 




HENRY FROWDE, M.A. 

ruvusuEK TO THE umvEAsmr or oxford 

LONDON, EDINBUB6H 
NEW YORK AND TORONTO 



lU 



CONTENTS 

PART m 

The Sea 

PAGE 

Chapter I. Conflict of Opinion regarding the Displacement of the Strand : 
Termiiiology and General Observationa Strabo. Dante. Celsias and 
Tiiimaeng. Nordenankar. Flay&ir and Leopold von Bnch. Goethe. Lyell and 
Darwin sua rapporters of the theory of elevation. Bravais and Engine Robert. 
Chambers and Domeyko; renewed expressions of donbt Theories based on 
gravitation. Adhemar and his successors. Summaries; Howorth. Terminology. 
Gennine case of dislocation in Mew Zealand. Plan of the foUowing chapters . 1 

Chapter n. The Ontlines of the Atlantic Ocean. The Canadian shield. The 
Baltic shield. Glint lines. The table-land of Spitsbergen. Greenland* The 
Caledonian mountains. The Armorican mountains. The Variscan mountains. 
The syntazis of Central Europe. The Iberian Meseta. Survey of the pre- 
Permian mountains in Europe. The islands of Europe. Western Africa. The 
east of Central and South America. Survey of the outlines of the Atlantic • 30 

Chapter IIL The Borders of the Pacific Ocean. New Zealand. Australia. 
New Caledonia. The sea of Banda. Borneo. Cochin China, Tongking. The 
Philippines. Formosa and the Liu-Eiu islands. Japan. The Kuriles and 
Kamchatka. Generalsurveyof the island arcs. Eastern China. North-eastern 
Asia. The arc of the Aleutian islands. The west coast of America . . • 143 

Chapter IV. Comparison of the Atlantic and Pacific Ontlines. The Atlantic 
structure. The Pacific structure. Their dissimilarity. Distribution of islands 
and volcanoa Overthmsting of the depressions. Progressive completion of the 
Mesozoic series towards the Pacific coast 201 

Chapter V. Palaeoxoic Seas. Introduction : The abyssal region ; cycles ; thick- 
ness of the sediments. The North Atlantic continent. Upper limit of the 
Sflurian. Universality of the middle Devonian transgression. The Carboniferous 
system. Paialic beds. Alternation of coal beds and marine sediments. Trans- 
gression of the Carboniferous limestone. The Permian system. Gondw&na 
land. General Survey 206 

Chapter VI. Mesoxoic Seas. The seas of the Trias. Positive movements in 
the Rhaetic epoch. Continuation of positive movements during the Jurassic 
epoch. Negative phase in central Europe and commencement of the Cre- 
taceous epoch. Later transgressions and intermixture of the Cretaceous faunas. 
The Cenomanian transgression. General survey of the Hesozoic seas • . 256 

Chapter Vn. Tertiary Seas and recent Limestone Formations. Negative 
phase at the close of the Cretaceous epoch. The central Mediterranean of the 
Terfeiaiy aera. The east coast of North America. The Tertiary region of 
Patagonia. Becent limestone formations. Summaiy 296 

Chapter Vm. The Strand-llnes of Norway. From Tjoalma Yagge to the sea. 
From lake Tome& to the sea. Movement of the ice contraiy to the slope of the 
valleys. Origin of the glint lakes of Lapland. Ancient strand-lines of the iQords. 
Origin of the seter. Comparison with Greenland. Vestiges left by the retreat- 
ing ice 826 



I 



iv CONTENTS 

PAGE 

Chapter IX. The Temple of Serapis near PuzzuolL The north-west coast of 
Italy. Situation of the temple of Serapis in the Phlegraean crater. The temple 
np to the year 1538. The eruption of 1538. Excavation of the temple and its 
present condition. Varions attempts at explanation. Volcanic phenomena . 364 

Chapter X. The Baltic and the North Sea during the Historic Period. 

Salinity within the Skager Rack. Mean level of the Baltic on the German 
coast. Oscillations on the coasts of Sweden and Finland. G-eneral survey of 
the negative displacement. Submerged forests and peat bogs of the North 
sea. The haffs and peat bogs of the Baltic coasts 393 

Chapter XL The Mediterranean dnring the Historic Period. The sea of Azov 
and the Black sea. Locus of maximum depression in the surfiu^e of the Mediter- 
ranean. The western Mediterranean. Venice. The Dinaro-Tauric region. 
The south-east Mediterranean. Conclusion * . 431 

Chapter XIL Strand*lines of the NortlL Diversity in form of the snrfiEice of the 
Oceans. Western coasts of the North Atlantic. Eastern coasts of the North 
Atlantic. The north of Eurasia and the west coasts of the North Pacific. East 
coasts of the North Pacific . . . 466 

Chapter Xm. Strand-lines of the Equatorial and Southern Coasts. Western 
coasts of the Atlantic Ocean, central and southern part Eastern coasts of the 
Atlantic Ocean, African part. Coasts of East Africa and Arabia. Coasts of 
India and Further India. Coasts of the Polynesian islands and Australia. West 
coast of South America 498 

Chapter XIV. The Oceans. The boundaries of the Oceans. Eustatic negative 
movement. Transgressions. Eustatic positive movement. Inadequacy of 
eustatic movements. River mouths and river terracea General survey of 
comparatively recent strand-lines. Oscillations of the Oceans. Alternating 
equatorial phases. The continents a result of subsidence. No appreciable 
change of level in the historic period 585 



LIST OF ILLUSTRATIONS 

PLATES 

^ I («VII). AP^rtofTromBO-Amt Frontispiece 

^ II (bVIII). Diagrammatic Representation of the Buccessiye Eruptions 

in the Fhlegraean Fields To face p. 368 

>/ III («IX). The Monthly Mean Water-leyel for the years 1873, 1874 

and 1875 from ROnnsk&r to Nord Eoster ,,404 

FIO. PAGS 

1. The Earth and Ocean according to Brunetto Latini 5 

2. The Earth and Ocean according to Dante 6 

3. Hypothetical Section across the River St Lawrence and the Isle of Orleans . 35 

4. Arctic North America 40 

5. Scania 47 

6. Tiew from the Omesvarre, South of the Bals Fjord 58-59 

7. Tjoahna Vagge, Moraine Landscape on the Boundary between Norway and 

Sweden 61 

8. Spitzbeigen 68 

9. Diagrammatic Representation of the Structure of West Spitsbergen 70 

10. A General Sketch of the Structure of Scotland .78 

11. Diagrammatic Section of the Archaean Foundation and Fragments of Mesozoic 

Rocks on Mont LosK^re 112 

12. Section through the Anthracite Basin of Chalonnes on the Lower Loire . .113 
IS. Jurassic Limestone pinched into the Gneiss at Famigen in the Mienthal . .115 

14. The Basin of Asturias 125 

15. The Island of Antigua 186 

16. New Zealand 145 

17. Japan 178 

18. C^stisoma NeptunuSt Guer.-^iineY 211 

19. Petalophthalmua armiger. Will. Suhm 212 

20. PolyeheUs crucifera, Will. Suhm 212 

21. C&noeephdlitee Sulzeri, Schloth 213 

22. Cyclopean Eyes 214 

28. Aeidaspua mira, Barr 214 

24. Trinudeusomatus^Si^m 215 

25. Section near Iowa City 238 

26. FobbQ Forests of the Carboniferous Period 244 

27. View from the Oberlahner, Ascent to the Funtensee Alp, Steinernes Meer . 261 

28. The Valley of the Blue Nile 275 

29. Limestone Beds in Sombrero 312 

80. Uvea, Loyalty Group 816 

31. Lifu, Loyalty Group 316 

32. Diagrammatic Sketch of the Left Side of the Upper SOrdal .... 835 

80LX«AJ II b 



vi LIST OF ILLUSTRATIONS 

FIO. PAGE 

38. Diagrammatic Section across the Left Side of the Upper SOrdal . . 886 

84. The Seven Sisters 387 

85. Closure of the Neu-Haufen Dyke, Schattau 843 

36. Dalager's Nunataks 844 

87. Frederikshaabs Isblink 358 

88. Jakobshavns IsQord 859 

39. Monte Olibano, Seaward Face 378 

40. The Volcanic Eruption of Monte Nuovo and the Shore abandoned by the Sea . 880 

41. The Mausoleum in the Bay of Makri 449 

42. The Third and Fourth Teeth of the Pettini di Ragusa 454 



PART III 

THE SEA 

' CHAPTER I 

CONFLICT OP OPINION REGARDING THE 
DISPLACEMENT OF THE STRAND : TERMINOLOGY 

AND GENERAL OBSERVATIONS 

Strabo. Dante. CelsiuB and Linnaeus. Nordenankar. Playfair and Leopold von Bach. 
Gk>ethe. Ljell and Darwin as supporters of the theory of elevation. Bravais and 
Eugene Robert. Chambers and Domejko; renewed expressions of doubt. Theories 
based on gravitation. Adh^mar and his successors* Summaries; Howorth. Terminology. 
Genuine case of dislocation in New 2iealand. Plan of the following chapters. 

We have descended from the mountains and stand on the seashore. The 
eye roams unchecked over the vast expanse of waters. A great wave 
approaches and seems about to reach us ; suddenly its crest curls over, it 
plunges downwards, and with a dull roar, sweeps a little further on*with- 
out wetting our feet. Then the water streams back, and a long green 
ridge of seaweed remains to mark the limit of its advance. Soon a second 
wave follows, then a third, and from time to time one somewhat higher 
than the rest, which whirls the seaweed further up the beach, and drives ufr 
back towards the foot of the cliff 

The roll of the waters is repeated like the chorus of a mighty song. 
The sublimity of the scene might hold us captive for whole hours. At- 
length the crest of the wave breaks at a greater distance from us, and soon 
the lowest point of the ebb will be reached. Then the Ocean returns and 
all the bands of white shells, and the green coils of seaweed it had left 
behind, are again swept together by the ever advancing line of the waves, 
until after six hours of rising tide even the spot where we first stood i» 
reached, and at last the sea washes once more against the foot of the cliff 

So to a stately measure the heavenly bodies cause the swaying Ocean to* 
alternately advance upon the land and retire* 

Let us now turn to the cliff Here the traces of an older strand may be 
eleariy seen, standing high above the existing level of the sea. For mile 
after mile they may be followed at a constant height, undisturbed by the 
nature or structure of the coast, over cUfb of limestone, granite, or ancient. 

tOLLAS n B 



2 DISPLACEMENT OF THE STRAND [partiii 

volcanic ash, or late Tertiary detritus ; they encircle as with a girdle, not 
only the mainland, but also the islands off its shores. 

This is something very different from the crushing and overthrusting 
we meet with in the mountains, and very different from folding, which is 
determined by the strength and direction of the tangential stress, by the 
rigidity of the rocks, and the resistance offered by opposing masses, changing 
. consequently in character as it passes from place to place. The phenomenon 
before us is of an altogether different nature, and if we recall the play of 
the tides with their rhythmic rise and fall in phases of half a day, then 
nature herself seems to suggest the question whether other forces may not 
exist capable of causing much more important oscillations and of much 
longer periods than that which we have pictured as driving us back to the 
foot of the cliff 

Many distinguished observers have answered this question in the 
affirmative ; others have supposed that changes may have occurred in the 
volume of water in the Ocean ; others, on the contrary,have imagined uniform 
and gradual movements of the solid Icuid — seculiir movements, as they are 
generally termed These conflicting theories have been advanced and 
adopted in turn according to the state of knowledge at the time or the 
views which happened to be in favour regarding the origin of mountain 
chains and the conditions of equilibrium of the sea-leveL 

The ancients were well aware that the sea once extended inland as far 
as the oasis of Jupiter Ammon, and that it covered all the low-lying ground 
from the Casius to the Red Sea : the polemic in Strabo's Geography (I, 3) 
bears striking witness to the zeal with which they sought to solve this 
problem. In accordance with Archimedes, Strabo asserted that the surface 
of all undisturbed water tends to correspond with that of a sphere, which 
has the same centre as the earth. Seas, he maintains, have no slope. He 
reproaches EIratosthenes with credulity for having believed on the evidence 
of certain engineers that the water on the two sides of the isthmus of 
Corinth stood at different levels. According to Strabo, not only islands 
and isolated mountains are subject to elevation, but also the mainland 
itself ; on the other hand tracts of land of greater or less extent are liable 
to give way and sink towards the interior. — 

Space would fail, were I to attempt to write the history of this discussion, 
which is as old as our science itself ; yet as previously we illustrated the 
nature of volcanos and the diverse origin of earthquakes by certain selected 
examples, so now we will attempt by means of a few chapters of this history 
to bring into relief some at least of the successive aspects which this ques- 
tion has assumed. 

Now, however, it is no longer to the mute eloquence of nature that we 
must lend an ear, but to the conflict of human opinion, sometimes loud- 
voiced enough. — 




CH. I] VINCENTIUS OF BEAUVAIS 3 

It is January 20, 1320. The bells of Verona ring in a bright Sunday 
morning, and the crowd greets with respect a personage of tall stature and 
earnest countenance, approaching with slightly inclined head to enter the 
ehapel of Santa Helena : it is Dante. 

All that can move the human soul the great poet has felt, and in the 
realm of imagination he has travelled farther than any mortal man before 
hijiL He has survived the loss of his Beatrice, and the loss of the emperor 
from whom he had hoped a better future for bis country. Now flying 
from the hatred of his native city he has found a refuge at the court of the 
leader of the Ghibellines of upper Italy, Can Grande of the house of the 
Scaligeri. With a gift of description never before equalled he has led his 
astonished contemporaries to the heights of the blessed and the depths of 
the lower world ; and now he returns to the starting-point of his mightiest 
creation, to the examination of that which is greater than all the concep- 
tions of poetry, the actual constitution of the Universe. 

Dante has to-day invited the whole educated world of Verona to listen 
to a discourse in this chapel, entitled * De aqua et terra.' He proposes to dis- 
coBB the relative position of land and sea, and as he tells us himself, every 
one came at his invitation, ' with the exception of a few, who feared by their 
presence to confirm the exceptional importance of others \' 

But it is difficult for us, accepting as it were this invitation after an 
interval of more than five hundred years, to examine without prejudice the 
views of the great poet. It is not easy to free ourselves from that presump- 
tion of superiority with which we are wont to regard the efforts of a period 
ao remote, in which nevertheless, with the assistance of Arabian science and 
the somewhat fragmentary remains of Aristotle, the foundations of the 
existing edifice were laid : a task accomplished, in spite of the paucity of 
observations, by the courageous and untiring efforts of the more eminent 
spirits to obtain a victorious comprehension of the Cosmos as a whole. 

The work of recognized authority in Dante's time was the ' Speculum 
quadruplex ' of the Dominican, Vincentivs of Beauvaisy completed in the 
year 1244. This monk, one of the ornaments of the court of Saint Louis, 
arranged the material in the oosmographical part of his encyclopaedia in 
accordance with the seven days of creation K Vincentius points out the 
contrast which exists, in the genesis of the elements, between the fiat lux 
on the one hand, and the congregenJtur aqua£ ut appareat arida on the 
other. In the expression congregeiiiv/r aquae he sees a condensation of the 

' Qoirtioiie trattata in Verona da Dante Alighieri 11 di 20. gennajo xoccxx intomo 
alia forma del globo terracqueo ed al Inogo respettivamente occnpato dall* acqua 
e dalla tena ; ed. Torri, Livorno, 1843, p. zlii, § zxiv. 

* Bibliotheca Mandi, Ben vener. viri Vincentii Burgnndi ex ord. praedicat. episcop. 
bellovac. Specnlnm quadruplex, naturale, doctrinale, morale, hittoriale ; fol. BeHerus, 
Doaci, 1624; lib. V, coL 307, et seq. 

B 2 



4 DISPLACEMENT OF THE STRAND [part in 

water vapour in the lower layers of the atmosphere, an accumulation of the 
resulting water in the hollows of the earth's surface, the awakening of the 
via iTiclinaiiva ad dedcendum, or, as we should say, of the force of gravity, 
previously slumbering in the molecules of water. As the blood communi- 
cates with the heart, so all the waters flow into the sea. The surface of 
the earth is spherical, so Ukewise is that of the waters. 

Thus far the conceptions of Vincentius are clear, and there is no funda- 
mental difference of opinion among his successors. In continuing his 
examination of the curvature of the Ocean, Vincentius is led to reflect on 
the rounding of the surface of water in a goblet, and the spherical form of 
a drop. The fact that the shore may be kept longer in view from the mast 
of a ship than from the deck, seems to him a proof of the independent con- 
vexity of the surface of the sea, and from the presence of springs in the 
high mountains, he concludes that the Ocean must stand at a higher level 
than the mainland. 

Thus Vincentius furnishes us with the first, though vague suggestion, of 
a force of cohesion, as seen in the law by which a drop is formed : this 
influences and indeed determines the form of the surface of the Ocean, which 
consequently does not depend solely on the action of gravity. We might 
almost say that he represents the Ocean to us as a single gigantic drop 
adhering to the globe ^. 

Other investigators, as Soger Bacon, in his remarkable 'Opus Maius,' 
which he presented to Pope Clement IV in 1267, start from the unity of 
the Centrum Tnundi and the concentric arrangement of all the spheres of 
the Universe : this view received its clearest expression, so far as it refers 
to the Ocean, in that famous chapter, where Bacon shoyrs that a goblet 
would be capable of containing more liquor in the cellar than in the tap- 
room. Every point in the surface of a liquid, says Bacon, lies equally 
distant from the centre of the earth, every such surface is thus a part of 
a spherical surface ; the nearer the centre, the smaller the radius of curva- 
ture, the more considerable therefore its elevation above the edge of the 
vessel \ 

Nevertheless the idea gained ground that the oceanic sphere possessed 
an independent form, its elevation above the mainland was held to be an 
incontestable fact, perceptible to the senses, and many authors spoke of it 
as rising in a definite asymmetrical ridge. Among these was Brunetto 
Latini, that teacher, who had instructed Dante * how man may inmiortalize 

^ Op. cit. lib. VI, col. 877, cap. xiL ' Quod etiam OceanuB terrain cingens in vertioem 
sit coactuB.' W. Schmidt justly lays emphasis on the importance of this passage, and on 
the contrast between aelf-cofifarmation Kadflgure of the earth ; see his dissertation, Ueber 
Dante*s Stellung in der Geschichte der Eosmographie, 8vo, Graz, 1876, p. 10, note. 

* Fratris Rogeri Bacon, ord. minor., Opus Maius adi Clementem lY, Pont. Rom. ; ed. 
Jebb, fol., Lond., 1788, p. 97, cap. x. 



CH. i] DANTE 5 

Mmaelf ^' After the defeat inflicted by Manfred on the Florentines neu- 
H<Hite AperU in 1260, Bronetto, who was also a diBtinguiahed Btatesman, 
retired into exile in France, where he remained until 1267, i. e. until after 
Man&ed's death at the battle of Beoevento. It waa during this period that 
be composed his chief work ' Li IJvres doa Treeor' : in this aleo the theory 
of the elevation of the sea above the land was maintained, chi^y on account 
of the elevated situation of many springs '. But in Dante's case the intoi- 
tion of a scientific mind gradually entered into conflict with the views of 
his beloved teacher, and that is probably the reason why Dante in the whole 
of his discourse 'De aqua et 
terra 'does not once mention 

the name of any of his JO^ 

opponents \ ^ 

Dante b^ns by enu- ' 

merating the ai^^uments of 
the rival schools, and shows 
that they lead to two alter- 
natives; either the whole ' 
masB of water is situated 
eccentrically or there is a 
local swelling ap of part of 
it. He asserts Uiat neither _uj 
of these hypotheses is ten- 
able. If the mass of the 
Ocean were eccentric, then 
water must be ci^ble of 
Sowing both upwards and 
downwards. Let the points A and B (Fig. 2) represent the two centres. A mass 
of earth and a quantity of water filing from Z to the earth would proceed 
in different directions : but gravity is a property common to all bodies. 

For the same reaaon a local protuberance of the Ocean is aleo impossible, 
[or it wonld be at once dispersed. 

The Ocean is thus concentric with the earth and every part of ita surface 
is equidistant from the common centre. Since, however, the shores of the 
continents and the continents themselves rise above the surface of the 

■ Inferno, casL XT, t. 82-85. 

* Bnmetto lAtini, Iii Livres dou Tresor, pubL par P. Chabaille, Collection de docn- 
■leDta in^tt anr lliistoire de France, pabliie par lea soint du Ministre de I'lnitmctioD 
Pnbliqne, 4to, Paria, 18S3. In paxtionlar liv. T, cap. cvi, pp. 114-116; also p. 169 : ' et 
Hanritaine fenit en hante mer de Egjpte ; et commence cele de Libe, ou il ; a trop 
Sere* merreilleB ; oar la mer i e«t awei pint haute qae la terre, et le retient dedani sen 
uargei en tele maniere que ele ne chiet ne ne deconrt aor la tetre.' 

' Inferno, cant. XT, v. 119, 120: 'Siati laocomandato il mio Tesoro Nel quale io 



DISPLACEMENT OF THE STRAND [part m 



water, these parts of the earth must be higher than the Ocean. Again, it 
is evident that the earth rises from the Ocean owing to particular elevations 
of its mass, and not as the result of eccentricity, since in the latter case the 
dry land would be bounded by a circular outline, and this we know is not 
the case. 

There exists then a true elevation of the land, and of this we must 
seek to discover the cause. ' Dico igitur^' says Dante, * q^uod causa huiua 
elevationis ejfficiens non potest esse terra ipse; quia qv/um devari sit 
quoddarfi ferri sursum ; et ferri sursu/ni sit contra noituram terrae *.' 
The earth cannot elevate itself; nor can the cause be water, fire, or air ;. 
the elevating force must therefore be sought in the heavens. 

In the starry firmament this force 
cannot for obvious reasons emanate 
from the moving stars, the moon,, 
sun, or planets ; it must therefore be 
sought in the fixed stars, which exer- 
cise this influence, either after the 
manner of magnets or by the pro- 
duction of active vapours. Dante 
thus adopts the system which had 
already been expounded in greater 
detail by Ristoro cPArezzo in 1282 ; 
according to this not only must 
the irregularities of the earth's 
surface be ascribed to the fixed 
stars, but mountains and valleys 
present as it were a mirror of the 
various distances of these stars 
from the earth, in an inverted sense, 
like the impression of a seal in wax K 
In conclusion, Dante reminds his hearers that the apparent sinking of 
the coast observed by the departing sailor simply results from the general 
curvature of the liquid sphere, and adds it has long been known that 
springs are not fed by water directly ascending, but are formed by the 
condensation of water vapour on the mountains ^ — 

^ Dante, Quistione, &c, § zx. De causa efficiente eleyationiB Terrae ; ed. Torri, 
p. xxxvi. 

' Bistoro d*Areizo, La Composizione del mondo, terto italiano del 1282 pabblicato da 
E. Narducci, 8vo, Roma, 1859, p. 79. In addition Ristoro is aware of the erosive power 
of water, and of the existence of fossilized remains of marine animals : he even suggests 
the influence of earthquakes on mountain structure, p. 83, et seq. 

^ This was the opinion of Dante, the coemogsapher ; let us compare it with [the views 
expressed in] his poetry. Yincentius of Beauvais dedicates a whole chapter (lib. VI, 
cap. vii) to the question : * Quorsum iniectus lapis erit casurus, si perforatus sit terrae 




• Fig. 2. Copied from Dante: Qmestio d4 
forma et aUu aquae et ttrrae^ editio Torri, 
p. xxiv. 

A CSentre of the Earth and Heavens, 

B Centre of the Ocean ; 

t indicates the Earth, ft the Ocean, fff the 
Heavens. 



CH. I] DE MAILLET. SWEDENBOBG 7 

Dante died in the following year, a.d. 1821. Let us hasten through 
ihe centimes towards a period when as the result of increased observation 
that varied and almost uninterrupted succession of diverse theories arose, 
which has not even yet reached an end. 

In 1692, a young nobleman of Lorraine, Berioid de Maillet, was 
iq>pointed French consul in Egypt; transferred in 1708 to Leghorn, he 
was afterwards sent on several occasions to the Levant, and thus became 
acquainted with the coasts of the Mediterranean. He observed indications 
of the retreat of its waters, and concluded that the volume of the Ocean 
suffers a continuous diminution: his results were not published till after 
his death in 1748 ^ 

Meanwhile attention was being directed in Sweden to facts of the same 
kind In 1702, the physicist Hjcunie, having discovered that the coast-line 
was receding, caused marks to be cut in several of the rocks in order to 
measure the rate of the movement. Emmcmtid Swedenbarg believed the 
sea to be sinking, and this to a greater extent in the north than in ihe 
south : in proof of this he called attention to the rapid increase in the area 
of Lapland. The idea of a change in the form of the liquid envelope of 
the planet here finds definite expression. In a letter of May 21, 1721, 
addressed to Jacobus a Melle (von Honig) of LUbeck, Swedenborg asserts, 
not indeed as a fact, but as a hypothesis worthy of consideration, that ' the 
water of the Ocean sinks towards the poles and (probably) rises towards 
the equator, and that islands previously separated from the land have been 
united with it by the sinking of the sea^' 

Swedenborg took no part in the discussion of this question which arose 
a few years later in his own country : his name is scarcely mentioned in 
connexion with it. The poet Dante, after he had achieved the greatest 
creations of his genius, was still capable of high endeavour in the pursuit 

giobxts.* It will rest in the centre, Bays Yinoentins. It appears to many authors of that 
time to be an ahnoat insoluble problem that an element so light as fire should appear in 
the interior of the earth, since the spheres are otherwise arranged according to the 
weight of their matter. Let us consider in addition the conception of Jerusalem as the 
centre of the inhabitable world, and the relations of the central fire of the earth to hell. 
Lucifer, caet down headlong, reaches the earth at a point exactly opposite to Jerusalem, 
and travels to the centre of the globe ; the principle of evU is now identified tcith the force 
ofgrofoity. Ct the sketch of the centre of the earth by Philalethes, Uebersetzung von 
Dante's GcttUcher KomOdie, 8. Aufi., 1877, Taf. II, III. 

* TeUiamed, on Entretiens d'un philosophe indien sur la diminution de la mer, ave^ 
on miswonnaire fran9ai8 ; vol. 1, 1748, toL II, 1755. Some absurdities and a certain lack 
of ciitaeism as regards the &cts enumerated may have contributed to cause this work, 
in which e. g. Bronn*s Terripetal theory is clearly anticipated, to fall too early into 
oblivion. Qnatrefages has done it justice in his 'Charles Darwin et ses precurseurse 
fran^aifl,' 8vo, 1870, p. 19-82. 

' Epistola nobiliss. Emanuelis Swedenboigii ad Vir. celeberr. lacobum ^ Melle. 
Ada literar. Snec. 1, Upsala et Stockholm, 1721, p. 196 ; also by the same, Miscellaneae 
observationes circa res naturales, 8vo, Leipzig, 1722, 1, p. 47. 



8 DISPLACEMENT OP THE STRAKD [part ni 

of physical truth: the naturalist Swedenborg on the other hand after 
a long scientific career turned his attention towards the spiritual world. 
The poet may attempt to cross the limits of the earthly, he is lord in the 
kingdom of dreams; but if the man of science, whose heavier foot is 
accustomed to the firm ground of proof, venture into this realm, then while 
he thinks he is perceiving realities visions are his masters. 

In the region of the Mediterranean, two distinguished hydraulic 
engineers, Manfredi of Bologna and Zendrini of Venice, had meanwhile 
made observations, which led to results entirely opposed to those of 
de Maillet, but in harmony with the suggestion of Swedenborg. In their 
opinion the surface of the Mediterranean, so far from being in process of 
sinking, was slightly rising. The discovery of a marble pavement beneath 
the cathedral of Ravenna, eight inches below high-water mark, and the 
constant flooding of some of the lower parts of the Doge's palace in Venice, 
as well as the condition of the church of Saint Mark, were the evidence of 
this movement. Manf redi's treatise, * De aucta maris altitudine,' did not 
appear till after his death in 1746 : in this he explained the rising of the 
sea-level as due to the great quantity of sediment yearly carried down to 
the sea by rivers ^ 

Let us now return to Sweden. 

It is the autumn of the year 1729. In the university garden of Upsala 
sits a poorly dad student of medicine and jots down the names of the 
plants. His poverty is such, that, as he confessed later, he cannot afford 
to have his shoes mended, and is obliged to put paper in them that he may 
not walk on the bare sole. This is Linnaefwa, son of the vicar of Sten- 
brohult, now in his twenty-second year. An ecclesiastic of some note, 
Olaf Cdsivs, subsequently provost of the chapter of Upsala, and now 
engaged in preparing a history of plants mentioned in the Bible, addresses 
the student, questions him concerning a number of botanical names used 
by Toumefort, is delighted with his knowledge and takes him back to 
his house. The young man's zeal soon wins for him the full favour of 
Celsius and the intimate friendship of his nephew Andreas '. 

^ £. Manfredi, De aucta maris altitudine ; De Bonon. Scient. et Art. Institute atq. 
Acad. Gommentarii ; torn. I, pars 2, Bonon., 1746, pp. 1-19. Manfredi had made his 
obserrations in 1731, and died not long after, in 1739 ; cf. op. cit. tom. II, pars 1, 1745, 
p. 237 et seq. Ravenna, according to the testimony of Vitruvius, was moreover entirely 
built on piles; to what extent even in early times its low situation gave rise to 
complaints is shown by the passage quoted by Manfredi from Sidonius Apollinarius, an 
author of the end of the fifth century, who calls Ravenna a marsh where, contrary to aU 
the laws of nature, 'muri cadunt, aquae stant, turres fluunt, naves sedent, aegri 
deambulant, medici iacent, algent balnea, domicilia confiagrant, sitiunt vivi, natant 
sepulti* The graves thus extended downwards into the water. 

' Afzelius, Linn^'s eigenhftndige Aufzeichnungen tiber sich selbst, aus dem Schwe* 
dischen v. Lappe, 8vo, Berlin, 1826, p. 12. 



CH. I] CELSIUS AND LINNAEUS 9 

It was in 1724, while still a youth, that Andreas Celsius had visited the 
shores of the golf of Bothnia. He had observed the retreat of the sea at 
Hnddiksvall, Pitei, and Lulei. At Tomei he had been shown to his 
astonishment that the harbour constructed in 1620 was already useless. 
Mariners pointed out places, now scarcely navigable for boats, on which 
great ships had once sailed. Near Langelo and in other places he saw 
rings to which ships had once been made fast, now at a distance from the 
sea. From these facts he inferred a sinking of the sea-level to the extent 
of about forty-five inches (133 cm.) in the century. 

Let us now pass over the fourteen years which followed the meeting 
in the botanic garden. Linnaeus and Andreas Celsius have both become 
professors in ordinary at the University of Upsala. They have travelled 
much in Europe, and Celsius has enjoyed the friendship of Manfredi in 
Bologna; he has also taken a prominent part in the measurement of 
d^rees made by the French in the north ; they have each visited Lapland ; 
each enjoys the highest reputation in the learned world. They agree 
tc^ther to put forward in University lectures the much discussed and 
much opposed views of Celsius on the shrinking of the sea \ 

Linnaeus delivered his lecture on April 12, 1743, Celsius followed on 
the 22nd of June in the same year : in giving an account of these discourses 
I will take first that of Celsius, since he is the greater authority on this 
subject, and preaente us on this occasion with a summaiy of his views. 

Celsius adopts the following line of argument : — 

We must consider our planet in three successive states: the first is 
tiiat of inundation, the last that of conflagration, and betw^n these is an 
intermediate state. As regards the state of inundation, which commenced 
very long ago, we have information of various kinds. We now live in the 
intermediate state, but everywhere we may observe signs of the retreat of 
the sea; its volume is diminishing, and so we are hastening to the last 
state, that of conflagration: the sun has already entered upon this, the 
planets are in various stages of the intermediate state, more or less removed 
from inundation and conflagration. 

Linnaeus adds to this, as a kind of supplement, the following : — 

The Lord did not create many individuals nor were they scattered over 
the whole world-<to what end the creation of many when the same object 
may be attained by means of OTie pair or even one individual ? Year by 
year the sea recedes, and the outlines of the mainland are correspondingly 

' C. Linnaei Oiatio de Tellnris habitabilis Incremento, et A. Celsii Oraiio de Mnia- 
tioiiibas generalioribiu qnae id Saperficie Coiporum Coelestium contingunt, 8vo, Lugd. 
Batay.y 1744 ; for the actual facts also : A. Celsius, Anmerkung von Verminderung des 
Waawn in der Ostsee nnd dem Ostlichen Meere; der kOn. scbwed. Akad. d. Wiss. 
Abhandl., Ac., auf das Jahr 1748, translated by A. G. Eftstner; V, Hambuig, 1751, 
pp. 25-37. 



10 DISPLACEMENT OF THE STRAND [pabtiii 

extended. Originally the mainland was only a little island oontaining 
everything which the Creator had appointed for the use of the homan 
race. This island, that is to say Paradise, we may imagine as a lofty 
mountain under the equator ; over the slopes, from foot to snow-covered 
summit, every species of plant found the climate suited to it. As the 
mainland increased in size the seeds of these plants, often of peculiar and 
specially adapted structure, were distributed by winds, rivers, and birds. 

Andreas Celsius died in 1744, the very year in which these two lectures 
appeared in print in Leyden. The conclusions of Linnaeus were on all 
sides favourably received, owing to their conformity with biblical tradition, 
whereas those of Celsius, although the outcome of the same general view, 
called forth the most violent opposition, particularly his reference to the 
approaching destruction of all living things. Affairs reached such a point, 
that when the state historian, Olaua DaZin, brought forward fresh proofs 
from historical sources in favour of Celsius, the Swedish diet expressed its 
disapproval of his theory by a special decree. 

This, however, did not temper the keenness of the controversy. Many 
supported the views of Celsius ; some attempted to explain the shrinking 
of the Ocean by evaporation into space, others returned to an older theory 
of Pontoppidan, revived of late years by Saemann, according to which the 
fixation of water vapour had taken place during the consolidation of the 
globe. On the other side, the leading opponent of the theory was Bro- 
wallivs, the learned bishop of Abo. He pointed out that the low island of 
Saltholm, near Copenhagen, was in existence as far back as 1280, and 
consequently that the observations made by Celsius further to the north 
were- not confirmed in the south : he showed also that in the north itself, 
on the coast of Finland, in the district of Abo and in Bjomeborg's Lan, 
trees of great age stand but a few feet above the actual sea-level ; as for 
instance in Bjemoskargaard, where a tree was felled which stood only four 
feet above the water-level although it was 810 years old. To these very 
important and indisputable facts I shall recur later ^ 

The work of Bishop Browallius made a deep impression, and treatises 
on this subject became less numerous : finally, in 1792, that is almost fifty 
years after the two discourses of Celsius and Linnaeus, Admiral Norde- 
Tiavkar, who possessed a thorough knowledge of the Baltic, advanced new 
views on the undoubted sinking of the water-level. The work in which 
these were published failed unfortunately to find the appreciation it 
deserved^. According to Nordenankar the Baltic must be classed with 

^ J. Brovallius, Bet&nkningar om Vattensminskning, 8vo, Stockholm, 1755 ; published 
in German nnder the title ' Historische und phyaikalische Unteraachung der vorgegebenen 
Verminderung des Waesers und VergrOaserung der Erde,' &c., translated by K. £. Klein, 
8yo, Stockhohn, 1756. 

* J. Nordenankar, Tal. om StrOmg&ngame i Oster-Sjan, h&Uet for kongl. Vetenikaps 



CH.I] FEISL BBEISLAK 11 

inland lakee, the universal characteristic of which is that they stand at 
a higher level than the Ocean, just as the Malar stands at a higher level 
than the Baltic For this reason the decrease of the water is less 
sorprising. It was not till communication was established between the 
Baltic and the North Sea through the Oreeund and the Belts that the 
shrinkage of the water began: but at what period this connexion was 
effected we do not know, nor do we know when the waters of the Baltic, 
either by steady eflux or by a sudden outburst, will be brought into 
equilibrium with the Ocean; and not till this happens will the sinking 
of the water-level come to an end. 

No fewer than two hundred rivers flow down from the surrounding 
land into the Baltic, and the water-level consequently varies with the 
year and the season. Steady winds accelerate discharge into the 
North Sea or check it, according to their direction, and thus affect the 
water-level. 

Nordenankar has thus the credit of calling attention for the first time 
to the peculiar conditions which prevail in the Baltic, and the important 
influence of the fresh water which flows into it. 

In Italy, meanwhile, a fresh approach had been made to the earlier 
views of Swedenborg. 

The first volume of the works of the mathematician Friai, published in 
Milan in 1782, contains a remarkable chapter 'De aucta et imminuta 
Marium Altitudine.' The author concludes from the observations of 
Celsius and Manfredi that the sea-level in the north is sinking, while 
in the Mediterranean it is rising. Distinguished investigators had, it is 
true, expressed the opinion that the mainland of Scandinavia ia being 
raised by the force of a subterranean fire (Buneberg), but the elevation 
of such vast districts and such mighty mountain chains cannot take place 
without great and prolonged concussions of the earth or without the 
subterranean fire somewhere making its appearance, and even then, such 
movements would be very unequal in extent. Frisi further observes that 
although all seas with free communication must adapt themselves to the 
fonn of some continuous curve, yet an increased velocity of rotation would 
produce subsidence of the sea-level towards the poles. Such an acceleration 
is experienced by all bodies in process of condensation, which rotate about 
a definite axis \ 

Even in the case of Italy doubts arose. Breidak examined the now 

Academien, yid Praeridii nedl&ggande, 18. Jan. 1792, Sto, Stockholm, 1792. My 
attention was drawn to this treatise, now very rare, by Dr. D. G. Nathorat and Dr. J. A. 
Ahlstrand, librarian in Stockholm. A German translation exists under the title ' Die 
StrOmnngen der Ostsee.* 

^ P. Frisii Opemm tomns I, Algebram et (^eometriam analyticam continens, 4to, 
Mediolani, 1782, pp. 270-270. 



12 DISPLACEMENT OF THE STRAND [PARrm 

classic example of the temple of Serapis at Pozzuoli ; he was obliged to 
admit that no satisfactory explanation had been found, and in the French 
edition of his Travels in Campania published in 1801 he suggests the 
theory that the land itself had sunk five meters and had then risen again 
to the same extent. The translator Paimm^Temly it is true, adds, 'This 
notion seems to be a jest ; it is like cutting the Qordian knot because we 
cannot unravel it^' 

The views of Frisi were opposed by Playfair in 1802, chiefly on the 
ground that coral reefs had been met with in the tropics below the existing 
sea-level; and returning to the older theory of Lassaro Moro, Playfair 
came to the conclusion that in Sweden the land was indeed in process 
of elevation. At the same time he emphasized the absence of trustworthy 
observations and the difficulty of obtaining them. The views we now 
regard as established, he remarks, are merely provisional and will be 
altered and corrected as knowledge advances*. 

Leopold von Buck soon followed with much greater decision in the 
same path. In the last days of September, 1807, he travelled from Tomea 
to the south. * It is certain,' he wrote then, * that the m/rfcuce of the sea 
cannot einJe; the equilibrium of the seas simply does not permit of it. 
Since, however, the phenomenon of diminution cannot be doubted, there 
remains, as far as we can see at present, no other course than to admit that 
the whole of Sweden is being slowly elevated, from Frederikshald to Abo 
and perhaps even to Saint Petersburg V ' 

A change of opinion was beginning to make itself felt The tlieory 
of desiccation^ as we will call the doctrine of de Maillet and Celsius, was 
gradually giving way to the theory of elevation advocated by Playfair 
and von Buch. 

We must not, however, imagine that the desiccation theory was at 
once abandoned. On the contrary, for a brief season it attained greater 
popularity than ever before. This was at the conclusion of the great wars, 
at the beginning of the last century, when millions of men were expecting 
the end of the world on July 18, 1816. Just as the individual, weakened 
physically by want of food or poverty of blood, is subject to mental 
disturbances in which depression, anxiety, and undefined presentiment of 
death struggle for expression, so the exhaustion of the people at that time 
mi^t be measured by the irresistible force with which the idea of the 
immediately impending end of all living things took posses8i<»i of whole 

' S. Breialak, Voyages physiques et lithologiqaes dans la Campanie ; tiadnit par le 
g^n^ral Pommereun; Paris, 1801, II, p. 170, note. 

^ The detailed treatise is contained in note xzi to Playfair's edition of J. Hntton's 
Tlieory of the Earth, Edinb., 1802. 

' L. V. Bach, Gesammelte Schriften, heransgeb. ▼. Ewald, Roth n. Eck, 11, p. 504. 
The ' Reise nach Norwegen und Lappland' appeared in 1800. 



CH. I] K. VON HOPF. GOETHE 13 

nations, as by an inteUectual epidemic. The newspapers took particular 
pleasure in discussing the theory of Celsius which, condemned years 
before as pessimistic, because it prophesied the destruction of all life, now 
suddenly reappeared to comfort tormented humanity, by the prospect 
of many thousands of years of existence. The dreaded date passed by; 
a little polemic followed, and then the whole incident dropped into 
oblivion '. 

The theory of elevation, however, had as yet by no means attained 
supremacy. Many of the most distinguished geologists of the time 
r^;aided it with disfavour. Cuvier and Brogniart, who had proved the 
repeated alternation of marine and fresh-water deposits in the neighbour- 
hood of Paris, nowhere speak in their works of elevation or subsidence 
of the land. The Cretaceous, so they say in substance, is deposited by 
a firfA ma ; this withdraws (66 retire), fresh-water deposits succeed, . . . 
another sea, populated by other MoUusca, returns (remerU) to retire once 
more {ae retire), &c.' The same expressions were employed by Omalius 
d'Halloy in 1813 ^ and later in 1827, when Constarit Pr4vost undertook to 
dispute before the French Academy the results of these stratigraphical 
investigations ; authoritative at that time, he too did not base his position 
on the theory of elevation. On the contrary, he doubted whether the 
presence of intercalated fresh-water formations could be regarded as an 
indication of the complete withdrawal of the sea, and attempted to explain 
the whole stratified succession around Paris simply by a repeated subsidence 
of the waters, thus returning to the fundamental idea of Celsius, now based 
on other arguments and clothed in a new form ^ 

Even in 1822, K. von Hoff himself, the conscientious critic who had 
devoted himself particularly to the investigation of such questions in 
Germany, was unable to adopt the theory of von Buch : he also believed 
that the sinking of the sea must necessarily be a universal and uniform 

^ Seel, Yom Weliunteigange, mit Bezieliung auf die verkOndete Wasserabnahme der 
Erde, Svo, Frankfurt, 1817 ; W ... n in the Mainzer Zeitong of the 15th Febmaiy, 1817, 
&c. Even at the beginning of this century discussions took place on the cause of the 
uniyersal diminution of the waters, which was considered to be fiilly established as 
the cause of these phenomena ; so for example, Poiret, Conjectures sur les causes de la 
diminution des eaux de la mer, Joum. de Phys., LX, an xni, pp. 226-237 ; Patrin, 
Bemarques sur la diminution de la mer et sur les iles de la mer du Sud, op. cit., pp. 806- 
323 ; Poiret, op. cit., in the next volume, pp. 17-22. 

' Cuvier et Brogniarfc, Description g^ologique des environs de Paris. The geological 
part was first published by Brogniart, 1808, in the Annales du Mus^, and then 
independently in 1811, later with Cuvier's ' Ossements fossiles.' 

* Omalius d'Halloy, Note sur le gisement du calcaire d*eau douce dans les d^parte- 
ments du Cher, de TAllier et de la Ni^vre ; Joum. de Phys., 1818, LXXYII, p. 104 
et pasnm. 

* Constant Prevost, Les continens actuels ont-ils ^t^, & plusieurs reprises, submerges 
par la mer? Compt. rend., lY, 1827, pp. 249-346. 



» 



14 DISPLACEMENT OF THE STRAND [pabt m 



\ 



phenomenon; von Buch'a 'bold idea' he regarded as *a truly desperate • 
means of explanation V ' 

This judgement won the full approbation of Qoethe, who in this con- 
nexion dedicated to von Hoff a remarkable essay on the temple of Serapis. 
' For what, after all/ wrote Qoethe^ ' is all this shoving up of the mountains ' 
but a mechanism which gives no aid to the understanding and no play to 
the imagination ? Mere words, without any corresponding ideas K' t 

'The earth cannot raise itself/ says Dante, 'that is contrary to its 
nature': there is a singular parallelism between these words and those 
of Wolfgang von Goethe written 500 years later. No one, in the history 
of the human intellect, stands above these two men ; indeed, few stand 
beside them. Their keen eyes perceived the difficulties, but even they were 
not able to find any satisfactory solution. The pressing need for explana- 
tion brought about a return to the theory of elevation. In the year 1834, 
von Hoff also conformed to the new theory of the gradual elevation of 
continents, yielding to the force of fresh data, and won over by the theory } 
of the elevation of volcanos and other mountains, which had meanwhile ' 
been further developed by Humboldt, von Buch, and their contemporaries. 

The theory of elevation soon received the most important support from 
the investigations of Charles Lyell and Charles Darwin. 

Charles Zyell travelled in the summer of 1834 to Sweden, satisfied 
himself of the truth of the alleged facts, and pointed out, even in his first 
accounts, that the evidence of a rising of the land is much clearer in the ' 
north than in the south ^ At a later period, he inclined to the view, based 
chiefly on Nilsson's observations, that the elevation was greatest in the ) 
north of Scandinavia, decreased towards the south, and disappeared near 
Sodertelje, a few miles to the south-west of Stockholm, but from this point 
subsidence set in and continued to the southern end of the peninsula, so 
that a tilting movement is thus in progress, and the limb south of the 
axis is much shorter than that to the north of it^ 

During the years 1832 to 1836, Charles Darwin accomplished his 
memorable voyage to the Pacific Ocean and to South America. The 
discovery of the structure of coral reefs led him at first to suppose that an 
extremely extensive subsidence was taking place in tropical regions over 
the greater part of the floor of the Pacific Ocean. Darwin attempted even 

^ E. E. A. von Hoff, Geschichte der durch Ueberlieferung nachgewiesenen nattlrlichen 
VerSuderangen der Erdoberflilche, I, 1821, p. 447 ; II, 1834, p. 816 etseq. 

* Wolfgang ▼. Goethe, Geologische I^obleme und Versuch ihrer AuflOsung. The 
essay on the temple of Serapis bears the title 'Axchitektonisch-natarhistorisches 
Problem.' 

' G. Lyell, On the Proofs of a gradual Rising of the Land in certain Parts of Sweden. 
The Bakerian Lecture, read Nov. 27, 1834, PkiL Tians., 1835. 

* e.g. in Principles of Geology, 11th ed., 1872, II, p. 190. 



CH. I] BRAVAia EUGENE BOBEBT 16 

to represent on a map the distribution of areas of elevation and subsidence 
within these regions over the whole world ^. 

A visit to the coast of South America, however, led to the discovery 
of a rising movement affecting the whole southern part of the continent 
down to about lat. 30** S., and this, judging from the successive terraces 
left by the sea, must in Darwin's opinion proceed intermittentty '. 

The wide distribution of this phenomenon in space and its intermittent 
character were the two facts which began to suggest fresh doubts. The 
principal argument in support of the theory of elevation was indeed the 
local variability in the clumge of level experienced by the strand, and now 
the range of the proofs began to exceed the limits of the premises. 

Bravais had asserted that two strand-lines in the Alten fjord, near 
Hammerf est^ were not horizontal, and that the slope of the higher terrace 
was greater than that of the lower; but this evidence could no longer 
be regarded as conclusive. £lie de Beaumont, in a detailed report, certainly 
thought he had succeeded In showing that Bravais' observations indicated 
some connexion between the rise of the Scandinavian chain and the emer- 
gence of the littoral terraces ; but the same report gives evidence of the 
distribution of similar terraces over the whole of northern Europe, and 
thus simultaneously supplies the refutation of this hypothesis. Further, 
it has since been shown that Bravais' observations were incorrect. The 
terraces near Hammerfest are just as closely parallel to the existing sea- 
level as everywhere else in Norway, and Bravais seems to have brought 
into the same line of measurement fragments of strand-lines and terraces 
which were really distinct \ 

The 'Commission scientifique du Nord,'of which Bravais was a member, 
also induded the geologists Durocher and Engine Robert. The report of 
Bravais, so completely in accord with the views prevalent at the time^- 
erroneous, none the less, as was subsequently shown — found recognition 
on all sides; while the complete and invaluable collation of the facts 

furnished by Eug^e Robert received scant attention. This work was laid 

• 

* G. Darwin, The Stractore and Dieiribntion of Coral Reefs, let ed., 8vo, 1842. 

* Id., G^logical Observationa on South America, Bvo, 1846, p. 26, et passim. 

' Rapport sur nn m^oire de M. A Bravais relaiif aux lignes d'ancien niveau de la 
mer dans le Finmark (M. ^e de Beaumont rapporteur); Gompt. rend., 1842, XV, 
pp. 817-849. In opposition to these views it will suffice to quote, among recent 
publications, T. Ejerulf, Einige Ghronometer der Geologic, aus dem Norwegischen 
tlbersetzt von R. Lebmann, Samml. gemeinverstftndl. Vortrftge v. Vircbow und Holtzen- 
dorff, XV. Serie, Hefb 352, 858, 1880, p. 14, and in particular K. Pettersen, Terrasser og 
gamle Strandlinjer, 3. bidrag, TromsO Museum's Aarsbefter, III, 1880, pp. 80-86 ; the 
same, translated by Lehmann in Zeitscbr. f. d. ges. Naturwiss., LIII, 1880, pp. 815-822. 
This treatise was probably not yet known to v. Decben when be spoke on this subject 
before the Niederrbeiniscbe Gesellscbaft fQr Natur- und Heilkimde at Bonn on 
November 8, 1880. 



16 DISPLACEMENT OF THE STRAND [pARrnr 

before the Acad^mie des Sciences in 1844^. It shows for the first time the 
extension of the phenomenon over the whole northern region, and although 
the author has nowhere ventured to draw the final conclusions from his 
observations, yet he evidently perceived the insufficiency of the prevailing 
theory. 

Eug^e Robert concludes his report with the following summary : — 

1. The vestiges of ancient strand-lines do not appear to be uniformly 
distributed over the globe ; in the southern hemisphere they seem to be 
comparatively rare. 

2. They appear to become more frequent as we approach the poles; 
but this may simply mean that they are better preserved in these regions, 
where a scanty population, lacking our great industrial resources, has 
effected scarcely any change in the land it occupies. 

8. They also become more sharply marked the more remote they are 
from the equator; but here again, as Robert suggests, an explanation may 
be found in the fact that the atmosphere and vegetation exert a less 
destructive action on the rocks in the arctic regions. 

4 The ancient sea-maigins for which the evidence is most conclusive 
attain the greatest height (162 to 195 meters) towards the north ^. 

This work of Eug^e Robert, which might have had a most stimulating 
influence, has been almost completely overlooked Nevertheless, the wide 
distribution of the terraces became increasingly apparent, and it was 
precisely those who dedicated themselves to the special investigation of 
this class of facts, who were least able to repress their doubts regarding 
the ancient theory of elevation. 

In 1848, Bobert Chambers published a very instructive work on ancient 
sea-margins, containing a comparative study of the terraces of North 
America, Great Britain, France, and Norway. The author lays stress on 
the fact that no case of elevation observed during the existing period has 
affected an area at all approaching in extent that indicated by these 
ancient terraces \ He remarks : — 

' The recession, accession, and second recession of waters indicated here, 
do not necessarily imply risings and fallings of our island, but may be 

' E. Bobert, Recueil d^olwervations on recherches g^ologiqnes, tendant ^ prouver, 
sinon que la mar a baiss^ et bajsse encore de nouveau sur ioat le globe, notamment dans 
I'H^misphere Nord, du moins que le ph^nom^ne de soul^vement, depois T^poque oil il 
a donn^ naisBance aux grandee chatnes de montagnes, n*a plus gu^re continue k ae 
manifesier que d*ane maniere lente et graduelle ; Compt. rend. 80 Juillet 1844, XIX, 
pp. 265-267. 

' Voyage de la ConmuBsLon scientifique du Nord en Scandinavie, en Laponie, au 
Spitzberg et aux FerOe pendant lea ann^es 1888, 1889 et 1840 sur la corvette 'La 
Recherche,* public sous la direction de Paul Ghdmard ; £. Robert, G^logie, vol. X, 
pp. 194-195. 

3 R. Chambers, Ancient Sea-Maigins, as Memorials of Changes in the relative Level of 
Land and Sea, 8vo, 1848, p. 820. 



CB.I] GHAMBEBS. DOMEYKO. DANA 17 






aeooimied for if we suppose some distant ocean bed sinking, then rising, 
I then sintiTig again. Perhaps it may be some such latent change which has 
produced thoee immersions of forests and those wearings of coasts, with 
which English geologists are familiar. I feel, at least, a particular difficulty 
in admitting partial subsidences of land in the British Islands, when I see 
sach uniform terraces. around their coasts, as, in that case, deflexions from 
the true lines ought to have been conspicuous, which I am sure they 
are not.' 

Bejecting for these reasons a local explanation, Chambers suggests that 
the extensiTe subsidence in the region of coral islands, that is, in the torrid 
zone, must cause the ocean waters to flow away from the poles. This is» 
so far as I know, the flrst attempt to bring into causal connexion the for* 
mation of atolls in the tropics and that of terraces in the higher latitudes. 

In the same year, 1848, a description of the terraces of the coast of 
Chili, by Domeyko, appeared. The author made a direct comparison 
between these and the terraces of Norway. Although, according to 
Bravais, the terraces of the Alten fjord are far from horizontal, and 
althou^ the lines marking the recent sojourn of the sea occur at very 
different heights, and at places so remote as C!oquimbo on the one hand 
and the Alten fjord on the other, yet in both these regions the number of 
Imes is very limited. We are led, therefore, to suppose that these phenomena 
are in no way dependent on local causes, but are connected with circum- 
stances which influence the great revolutions of the globe, and afiect both 
hemispheres simultaneously ^ 

The following year, 1849, Dana, having completed his great voyage in 
the northern part of the Pacific, expressed his belief that the rise is greatest 
towards the north pole : the opposite movement towards the equator K 

While the theory of elevation thus gained no support from a wider 
knowledge of the characters of these phenomena, a new school of thought 
arose, which inquired seriously into the question of the invariability of the 
oonditions which determine the equilibrium of the Ocean. Very different 
eaoaes have been suggested from time to time as capable of producing 
a universal alteration in the form of the oceanic surface: in all of them 
the force of gravity plays a leading part. I will therefore include all 
^ews of this kind — using an expression of James Croll's in a slightly 
extended sense — under the common name of gravitaiion theories. 

The controversy between Dante and his opponents turned on a question 
of gravitation. To the gravitation theories belong also the views of 
Swedenborg. At a later time, when Halley's theory was in favour* 
L Bernard supposed that the 'terrella,' which Halley conceived to be 

' Domeyko, M^moire rar le terrain tertiaare et lea lignes d'ancien niyeau de POc^aa 
da Sod, anz environs de Coqnimbo (Chili) ; Ann. des Mines, 1848, 4* s^r., XIY, pp. 15^162. 
* J. S. Dana in Wilkes, U.S. Ezplor. Ezped. 1849, X, pp. 670, 677. 

" 



18 DISPLACEMENT OP THE STRAND [PABTin 

moving independently within the hollow body of the earth, thus causing 
the displacement of the magnetic poles, was also responsil>le for changes 
in the form of the oceanic surface^. In 1804» Wrede, starting from the 
assumption that the centre of gravity of the globe is not necessarily 
coincident with its centre of figure, proceeds to show that the position of 
the former may be altered by the transport of sediments and various other 
causes. But this will bring about a change in the surface of the Ocean ^. 

These various views may all be included in .the group of gravitation 
theories, as well as those embodied in the important works of Adh^mar. 

Bertrand and Wrede sought the cause of the alteration within our 
planet or upon its surface, Adh^mar on the other hand outside it, in its 
relations to other members of the solar system ; his theory is thus based 
on considerations similar to those involved in the explanation of the tides. 

The work which laid the foundation of this theory appeared in 
1842 ^ Its main outlines are essentially as follows: — 

The inherent heat of the planet is scarcely appreciable at the surface, 
and may be regarded as constant The warmth of which we are sensible 
is almost exclusively derived from the sun. Any given place receives this 
heat during the day only, losing it again by radiation at night, so that 
with an equal duiation of day and night the amount of heat received 
during the day and that lost during the night balance each other. The 
length of the day. is thus one of the most important elements in the 
temperature of a place. At the south pole the hours of the night exceed 
those of the day by 168 in the year; it thus receives less sunshine, is 
consequently colder, and presents conditions far more favourable to the 
accumulation of ice than the regions around the north pole, where the hours 
of the day are more numerous by 168 than those of the night. This state 
of things is dependent on the position of the earth with regard to the sun, 
and on its movement Owing to the precession of the equinoxes, the 
equal duration of day and night in all latitudes would recur at the same 
point on the earth's orbit at intervals of 25,900 years, but since the 
simultaneous displacement of the perihelion must be taken into account, 
this period is reduced to about 21^000 years. While, in our hemisphere, the 
sum of spring and summer is now some days longer than the sum of autumn 
and winter, at the end of half this period, that is in 10,600 years, these 
relations will be reversed. In the year 1248 of our era, the equal duration 
of day and night coincided with the perihelion ; since that time the 

^ L. Bertrand, Benouvellements pModiques des continens terrestres, 8yo, an VIII, 
pp. 274-300. 

' E. F. Wrede, GreognostiBche Untersuchungen fiber die Bildbaltiflchen LAnder^ 
beaondeA tlber das untere Odeigebiet, nebst einer Betrachtung fiber die allm&hlige 
Verftndemng des Wasserstandes anf der nfirdlichen Halbkugel der Erde und deren 
physiftchen Uraachen ; 8vo, Berlin, 1804. 

' J. Adh^mar, R^volntionB de la mer, 8vo, Paris, 1842. 



CH^i] ADHfiMAR CROLL AND SCHMICK 19 

northem hemisphere has been gradually growing colder, the southern 
warmer. Up to the year 1248 this circumstance caused the continuous 
enlargement of the ice-cap which surrounds the south pole; by the for- 
mation of this cap the centre of gravity of the planet was displaced, and 
the coeans were drawn towards the south. This explains the greater 
expanse of ocean in the south, and the predominance of land towards 
tiie north. After 10,500 years, that is in about the year 11,748 of our 
er% the same state of maximum refrigeration and maximum submergence 
will be reached by the north pole. 

Thus the planetary movement would determine a periodic tran»fere7ice 
of the ice-capfrom one pole to the other, and an accompanying submergence 
of the oorresponding hemisphere. 

It must^ however, be observed that this conclusion is reftited by 
historic testimony. For if the northem hemisphere has actually been in 
process of cooling since the year 1248, and its ice-cap therefore in- 
creasing, then, since about one-seventeenth of the period preceding the 
mAximinn accumulation of the seas in the northem hemisphere has now 
already elapsed, some rising of the waters should be observed on all the 
northem shores, and this is not the case. Adh^mar certainly felt this 
contradiction, and made an attempt to meet it. He supposed that the 
Antarctic ice needed a very long period to disappear, an^ even suggested 
that a sudden adjustment might be possible as soon as the centre of gravity 
crossed the plane of the equator. The phenomena in the Baltic might 
perhaps be ascribed to local conditions. 

In spite of this and of many other weak points, Adh^mar's work was 
most stimulating in its influence as a serious attempt to explain, by a single 
and consistent theory, three great phenomena, namely, the predominance 
of water in the southern hemisphere, the periodic return of glacial epochs, 
and the universality and constancy of oscillations of strand-Une. Croll in 

■ 

England, Schmick in Germany, and many others, have amended and 
developed Adh^mar's views, all retaining the main idea of an accumulation 
of the sea, which is periodically transferred from one pole to the other. 
On the other hand, distinguished climatologists, and in particular A. WoeiJcof, 
have definitely asserted that the facts on which Adh^ar and his successors 
have based their theory, are incompetent to produce such extreme changes 
of climate as they have been supposed \ 

* J. Croll, Climate and Time in their geological relationi, a Theory of the secular 
changes of the Earth's climate, 8vo, London, 1875. This principal work had been 
preceded bj a number of smaller treatises which appeared, sinoe 1864, chiefly in the 
FhiL Mag. ; of those which followed I will only mention : Physical Causes of the 
Snbmeigence and Emergence of Continents, Geol. Mag., 1874, p. 809 ; Schmick, Pie 
Umsetznng der Meere und die Eiszeiten der Halbkugeln der Erde, ihre Ursochen und 
Periodeii, 8vo, Koln, 1869 ; by the same, Das Fluthphftnomen und sein Zusammenhang 
mit den sftcnlftren Schwankungen des Seespiegels, 8vo, Leipzig, 1874 ; by the same, Die 

C 2 



20 DISPLACEMENT OF THE STRAND [pabthi 

All explanations belonging to the group of gravitation theories pre- 
suppose that the sum of observations as to the oscillations in various parts 
of the world may, after the elimination of erroneous or doubtful cases, 
be expressed by some simple formula; in other words, that there exist 
great and continuous regions of changing level, the distribution of which, 
is related to the rotation axis of the globe according to some easily 
recognizable law. If, on the other hand, it can be shown that the distri- 
bution of these oscillations is sporadic, not following any recognizable law, 
then their origin cannot be sought in alterations of the form of the sea, 
the whole group of gravitation theories must be abandoned, and we must 
return, in spite of all objections, to the theory of movements of the solid crust. 

There is no lack of attempts to determine the geographical distribution 
of the so-called 'seculcur oscillations/ We may refer to the works of 
lUduz^, Peachd^, Hahn^, Issd^ and others; as well as to the little 
sketch-map by 0, R, Credner *. 

These attempts have not led to results sufficie^ntly uniform to come 
under the form of a law. All the authors just mentioned indicate eleva- 
tions and subsidences in the most diverse geographical latitudes, often, 
indeed, opposite movements in the most closely adjacent regions. 

The case is otherwise with the investigation made by H. H. Hovxyrth. 
In a number of tracts which have appeared since the year 1871, he has 
attempted to show, always from the standpoint of the elevation theory, 
that the land round the poles is rising, and he finally reaches the conclusion 
that an actual deformation of the planet is in progress, contraction taking 
place about the equatorial region, and proceeding thence in an increasing 
convexity, which probably attains its maximum over the magnetic poles \ 

Aralo-Easpi-Niederang nnd ihre Befunde im lichte der Lehie Ton den sftcolftren 
Schwankungen des Seespiegels und der W&rmezonen, 8vo, Leipzig, 1874. On the other 
side, in particular Pilar, £in Beitrag zur Frage fiber die Ursachen der Eiszeiten, Svo, 
Agram, 1876. As treatiseB all tending more or less in the direction of this branch of the 
gravitation theozy I maj also mention : Le Hon, Periodicity des ddnges, 8to, Broxelles, 
1858 ; Garret, Le d^placement polaire, prenves de la variation de Taze terrestre, 8vo, 
Paris, 1876 ; P^roche, Les ph^nom^nes glaciaires et torrides, causes auzquelles doivent 
6tre attribues la precession des equinoxes et les oscillations polaires, 8yo, Paris, 1877. 
On the other hand, A. Woeikof, Mitth. geogr. Ges. Wien, 1882, pp. 856-369, and Am. 
Joum. Science, 1886, XXXI, pp. 161-178. 

^ Reduz, La Terre, 3* ed., 1874, 1, pp. 709-767 ; general map, pi. xxiv ; also Revue des 
Deux Mondes, 1*' janv. 1865. 

' O. Peschel, Neue Probleme der vergleichenden Erdkunde, 2. Aufl., 1876, pp. 97-114. 

' F. G. Hahn, Untersuchungen tlher das Aufsteigen und Sinken der KQsten, 8vo, 
Leipzig, 1879. 

* A. lasel, Le oecillazioni lente del suolo o bradisismi, gr. 8vo, Geneva, 1883. 

' G. R. Credner, Die Deltas, ihre Moxphologie, geographische Terbreitnng und 
Entstehungsbedingungen ; Peterm. Geogr. Mittheil., Ergftnzungsbd. XII, 1878, pi. iii 

' Howorth's first statement that all the land round the north pole is rising, and ihe 
nearer it lie$ to thf poU the grrater the extent of the denxUion^ I find in Nature, Dec. 20, 



CH, 1] HOWOBTH. BELT 21 

« 

Strange to say, Howorth does not seem to have observed that his 
re0olt8> if they should prove true, would at once menace the veiy founda- 
tion of the elevation theory, and would raise the question whether this 
uniform result is not to be ascribed to an alteration in the form of the 
surface of the sea. It must be admitted, besides^ that Howorth's actual 
data are somewhat defective, and that many contradictory observations 
remain unexplained; on the other hand, we must not omit to mention 
that Thomas Bdt, following another method based on a study of the 
existing distribution of species, has arrived quite independently at a result 
which, in substance, exactly corresponds with that obtained by Howorth \ 
though it receives a veiy different explanation. 

Belt maintains that a rising of the waters about the equator has been 
taking place since the glacial period. The difference between the marine 
Mollusca on the two sides of the isthmus of Panama; the resemblance 
between the terrestrial Mollusca of the northern islands and the West Indies 
as far as Puerto Bico and those of Central America and Mexico, and the 
correepondence of the land shells of the southern islands partly with those 
of Venezuela and partly with those of Quiana; the well-known facts of 
geographical distribution presented by the Malay archipelago, and the 
gigantic statues of Easter island, are for Belt so many. proo& of this 
continuous rising. The author believes the explanation to lie in the 
periodic and simultaneous formation of ice-caps at both poles, as the result 
of changes in the obliquity of the ecliptic. 

The contraction of the globe about the equator imagined by Howorth, 
and the rising of the oceans about the equator supposed by Belt, are, 
however, only two different explanations of one and the same conception 
of the facts. 

This conception, however, is the same as that which long ago was 
held — though on different grounds — ^by Swedenborg and Frisi, and more 
recently by Robert Chambers. It presupposes on each side of the equator 
a nearly symmetrical and homologous arrangement of the regions of eleva- 
tion and depression, while from the premises of Adh^mar and his successors, 
the oscillations on the opposite sides of the equator must be opposed, 
Le. (xmiplementary. 

1871, pp. 162 and 163 ; mach more detailed evidence is given for the statement in 
a waA hj the same author. Recent Elevations of the Earth's Surface in the Northern 
Giicumpolar Regions, Joum. Geogr. Soc., 1878, vol. 43, pp. 240-263. The data for the 
Sooth Polar regions are contained in Nature, March 28, 1872, pp. 420-422, and the 
eoBclnsions drawn chiefly in Nature, Jan. 15, 1874, p. 201. Murphj maintains, op. cit. 
Jan. 18, 1872) p. 225, that the southern regions are also rising, so that the perimeter of 
the equator is diminishing ; cf. also Hamilton, op. cit., Jan. 25, 1872, p. 242, and Murphy, 
Feb. 8, 1872, p. 285. 

1 T. BeH, The NaturaHst in Nicaragua, 8vo, 1874, pp. 263-274. Also: The Glacial 
Peiiod in the Southern Hemisphere, Quart. Joum. of Science, July, 1877. 



22 DISPLACEMENT OP THE STRAND [pabt m 



ComtemjKMnuy litentore shows that opinions on this important 
question aie still widely divergent. 

Howorth and Belt, as we have jnat sem, were led by different paths 
to the same view, that which invcdves a synmietrical displacement of 
masses of water on each side of the equator. 

It may be shown that the opinion of many observers eminent in this 
branch of inquiry inclines towards the view represented by this last group 
of theories. I may mention, in particular, JiiZius von Boast \ the most 
competent authority on the ancient terraces in New Zealand, and Warren^ 
Uphafn^^ who has described the recent alluvial land in New Hamp- 
shire. In 1875, jAT. S, Shaler declared the theories of Adh^ar and Croll to 
be very improbable, since all observations are in favour of the simultaneous 
glaciation of both hemispheres, but he nevertheleaB expressed his conviction 
that it is the sea and not the land which is subject to movement \ 

The gravitation theory introduced by Adh^mar, which involves the 
asymmetric accumulation of the seas, has nevertheless, in the completed 
form which we owe to Croll, found numerous adherents in Elngland : it 
has been fully accepted by Charles Darwiny James OeUde, and many- 
others ; it is true that in most cases the theory has been applied to explaiti 
changes of climate rather than movements of the sea-leveL 

Charles Lydly whose long career produced such important results for 
our science, was always a keen opponent of the theory of the elevation of 
volcanic mountains, the theory, that is, of elevation craters, and an equally 
keen and influential champion of the theory of secular oscillations of the 
continents. On the formation of mountain chains, he never expressed 
himself with equal decision. In the later editions of his ' Principles,' all 
the older theories and arguments in favour of continental movements am 
retained ; but notwithstanding this, Croll's theory finds recognition as re* 
vealing a hitherto neglected vera causa of a certain oscillation of the 
sea-level ^ 

The elevation theory rests, even at the present day, on the alleged 
unequal movement of neighbouring regions, and the alleged tilting move- 
ment of whole countries such as Sweden and Greenland. In spite of the 

* J. y. Haast, Oeologj of the Provinces of Canterbniy and Westland, New Zealand, 8vo, 
1879, p. 381. 

* Warren Upham in Hitcbcoclc, Geology of New Hampshire, 8vo, 1878, III, p. 329 et seq. 

* N. S. Shaler, Notes on some of the Phenomena of Elevation and Subsidence of the 
Continents ; Proc. Boston Soc. Nat. Hist., 1875, XYE, pp. 288-292. A similar recognition 
of the return to the idea of the yanability of the ocean surface is also to be found in 
certain hjdrographic treatises, e.g. in Stahlberger, Ueber Seespiegelschwankungen, 
Mittheil. geogr. Ges. Wien, 1874, 2. Ser., TU, pp. 58-66. 

* Lyell, Principles of Geology, 11th ed., 1872, I, p. 279. The introduction to this 
edition is the best illustration of Lyell's position with regard to this question. Croll*8 
grayitation theoiy is recognized, but its influence on climate is not admitted to be so 
great as is claimed by other authors. 



CH. I] TRAUTSCHOLD. SUESS 23 

attacks of certain physicistfi, directed chiefly against the absence of a closer 
definition of the stupendous force, which is said to elevate and depress 
grettt parts of tiie earth's surface, the elevation theory has maintained its 
poaitHm as the aeoepted doctrine up to the present day, especially among 
geologists who devote themselves to stratign^hy ; from this doctrine the 
explanation of transgressions and gaps in the series of formations is 
diorived, just as it was sbcty or seventy years agoK 

Wrede's hypothesis that the position of the centre of gravity of the 
solid earth is altered by the displacement of sediments, was revived some 
time ago by G. JdgerK 

H. Trautechold has drawn his conclusions from a comparison of the 
nature and distribution of the ancient sediments ; in numerous treatises he 
has persistently maintained, in opposition to the prevailing opinion, that 
secular upward or downward movements of the continents do not take place \ 

This wide divergence of opinions on a question of vital importance in 
our science, has led me to submit all the actual observations hitherto 
obtained to a fresh examination. I felt the further incited to this task, 
because, relying on the teaching of revered masters, I myself made many 
years ago an attempt to bring the new views on the formation of moun- 
tains into harmony with the received doctrines ; I refer to my treatise en- 
titled ' Die Entetehung der Alpen.' I there maintained that the movement 
of Scandinavia might be attributed to the formation of a fold of great 
ampfitade within the earth's crust. Still, although I added that the wide 
distribution of certain marine deposits, especially of the middle Cretaceous, 
is not to be explained in this way, and, at the same time, expressed my 
belief that the temporary extension of the seas must depend on causes fiur 
more general, and probably subject to a certain degree of periodicity; yet, 
even with these qualifications, I could not conceal from myself that such 
an interpretation was unsatisfactory K 

The folding of the mountain chains and the formation of horizontal 
strand-lines which run without interruption over mountain segments of 

' 'Again, succeBnye strata of different kinds are accounted for by the sabsidence of 
land ; it sinks beneath the sea to receive its load of sediment, as the camel drops on its 
knees and then rises ; but, more patient and accommodating than the camel, it takes as 
many loads as the geologist is pleased to impose on it. This assumption of the 
unlimited Knking and rising of land is plausible and convenient, but it is inexplicable 
and unproved.* Besborough Cooley, Physical Geography, 8vo, London, 1876, p. 428. 
Siemens expresses similar views in Monatsber. Akad. d. Wiss. Berlin, 1878, p. 572. 

* G. J&ger, Die Polfltlchtigkeit des Landes, Ausland, 1865, p. 867, and 1867, p. 121. 

' H. Trautschold, Ueber s&culftre Hebungen und Senkungen, 1869 ; Sur Tin variability 
do mvean des men, 1879 ; Zur Frage fiber das Sinken des Meeresspiegels, 1880, and 
many later articles ; all in the Bull. Soc. Imp. des Natural, de Moscou. 

* Die Entstehung der Alpen, 1875, pp. 119, 150 ; Ueber die vermeintlichen s&cul&ren 
Schwankungen einselner Theile der Erdoberfi&che, Verhandl. geol. Reichsanst., 1880, 
pp. 171-180. 



24 DISPLACEMENT OF THE STRAND [pabt hi 

the most diverse description, are two entirely different things, and the 
elevation theory therefore ascribes to the lithosphere two movements of 
a completely different nature. An examination of the actual data on which 
this theory rests has proved^ however, to be an extremely difficult taslc 

In 1834, at the annual meeting of the Geological Society of London, 
not long after the publication of the last volume of Lyell's ' Principles of 
Geology/ the president Chreencmgh warned the members against too ready 
an adherence to the theory of elevation which was then rapidly gaining 
recognition. The statements regarding the elevation in Chili in 182J2 
were, he said, untrustworthy ; the term ' elevation ' was used in the most 
various senses, and of the uniform elevation of a whole continent it was 
almost impossible to form a conception; above all, a terminology was 
needed which should not involve any preconceived theory^. 

This warning appears to have produced no result, and it was not 
until 1848 that Robert Chambers introduced a new phrase; he spoke 
neither of elevation nor of subsidence, but only of 'changes of relative 
level,' or, as we shall say , displacements of the 8travd4ine. 

With the adoption of these neutral terms it at once follows that the 
displacements of the strand-line in an upward direction must be described 
as positive, those in a downward direction as negative, since this is the 
terminology universally employed by all oceanographers and in all 
operations of water gauging. Here we have no choice, and aU in- 
vestigators who have sought to abandon the adventurous sea of theories 
for the solid ground of fact, and have endeavoured to obtain information 
by direct measurements of the water-level from the shore, like Forssman 
for instance in his valuable investigation on the level of the Baltic, have 
adopted naturally the signs + and — in recording their results. In this 
work therefore the older term elevation of the land will be r^laced by 
negaiiive displacemevd of the 8tra7id4ine, and subsidence of the land by 
positive displacement of the drand-line. 

But if, equipped with this neutral terminology, we now attempt to 
proceed to a serious examination of the position, we find ourselves con- 
fronted with so many circumstances which may exert an influence on the 
sea-level, by so much uncertainty in the existing data, and by so many 
sources of error, that finally little remains as the result of many years' 
labour, but a conviction that many doctrines which in spite of the 
warnings of unprejudiced authorities have become accepted dogmas are 
erroneous, and a hope that the rising generation will succeed in obtaining 
a more exact knowledge of the laws which govern the statics of the seas. 
For this reason the following chapters in so far as they treat of displace- 
ment of the strand are mainly critical. 

' 6. B. Greenough, Address delivered at the Anniversaiy Meeting, Feb. 21, 1834 ; Proc 
GeoL Soc, 1833-1888, II, pp. 54 et seq. 



CH. I] METHODS 25 

There are three methods of studying changes in the sea-leveL 

The first consists in tracing the various extension of the andefnt 9eas. 
Although the strand-line itself can seldom be recognized, yet in the great 
transgrefisions, particularly that of the middle Cretaceous, we perceive the 
widely distributed signs of positive movement. Frequently, doubtless 
even as a rule, these transgressions are accompanied by erosion, and hence 
n^;ative movements are far more diflBcult to discover by this method than 
positive, and are only rarely brought to light, if indeed they do not 
frequently escape observation altogether. 

The second method is found in a comparative study of the Tudwre of 
the sedvmefnJtary formatume. Very distinguished investigators, Charles 
Darwin for example, have believed that sediments were only formed with 
a fdnkiTig sea-bottom, that is during a positive movement. This is by no 
means the case, but it is evident that with a predominance of the negative 
movement the approach of that critical moment is hastened when the 
upper surface of the sediment and the surface of the sea coincide at 
the same leveL For the present it is sufficient to refer to what has been 
said with r^ard to the superposed and apposed series on the isthmus of 
Suez (I, pp. 878 et seq.). 

The third method, finally, is the excvniination of the eadding coasty but it 
is precisely in this study of the evidence furnished by marks upon the 
coast that we meet with the greatest difficulties. Positive movements as 
a rule conceal their own traces, and it is only exceptionally, as in the case 
of coral reefs, for instance, that they are definitely recorded. Negative 
movements leave visible signs: but this is not all. Even when an 
oscillatory movement is taking place, in which the positive character 
predominates, it may often happen that the only evidence which remains 
is that of a negative movement. An example will make this clear. Let 
abj be, cd, represent equal spaces traversed in equal times, and let us 
suppose that we have a preponderance of the positive movement, such that 
the ratio of the positive displacement to the negative is as 4 : 3. We shall 
then have the following scheme : — 

m 



e 


+lm 


"TtlTl 


+ de 


-6/ +kl 


—no &c. 


-^cd 


--fg +ik 


—op +pq 


+ 6c 


--gh +Ai 


P 


+06 


h 


. 



a 

In 4 units of time the strand will traverse 4 positive units of distance 

from a to e, it will leave here its highest mark, and then sink through 

S negative units of time to A, it will then again ascend in 4 positive units 

of time to m, and so on. In this case, however, the relict strand-line at e 



26 DISPLACEMENT OP THE STRAND [PAKPin 

will remain visible through the 6 units of time^ included between ef and £2, 
of which 3 are negative and 3 positive, and it is only during the single 
unit of time Vm» that it will be concealed. Although the positive movement 
thus predominates in the ratio of 4:3^ yet the chances are as 6: 1 that 
a mark will be seen above the water-level from which negative movement 
may be inferred. Even in the case of a. double predominance of the 
positive movement, that is with a ratio of 6:8, there is still a twofold 
probability that a negative mark will be seen above the water-level. 

But in nature the movements do not take place either by regular 
oscillations or with regular intermittence ; and a whole series of factors 
of greater or less importance come into play and influence the position 
of the strand. The simple example given above is intended merely to 
show how necessary it is not to lose sight of compensating movements. 

In studying the strand we must further distinguish between the 
movements of the present day whidh may be directly recorded by tide 
gauges and those which can only be inferred from ancient remains such 
as the so-called ' raised beaches.' A link between the movements now in 
progress and those of the ancient world is afforded by evidence which lies 
within the limits of historic time, i. a to which a date or period can be 
assigned and approximately expressed in figures, as for instance in the 
case of the Lithodomus borings in works of Roman construction. 

It is obvious that the historic period, as we have defined it, is far ixoaa 
possessing the same importance and duration at the mouths of the Nile, of 
the Bhine, and the Mississippi ; but similar differences also exist in speaking 
of other periods, as, for instance, the glacial or pre-gladal. Greenland is 
still in the glacial period, and no one who has seen the moraines of Lapland 
in the 70th degree of north latitude will find it easy to believe that these 
are just as ancient as those of the moraine lands long abandoned by the ice 
at the foot of the Alps. 

One of the most dangerous sources of error lies, indeed, in the temptation 
to class together, without distinction, strand-lines of different age. Thus 
the oft-repeated assertion that Greenland is now undergoing a tilting 
movement rests solely on a mistaken association of extremely ancient 
high-lying strand-lines in the north with marks apparently positive on 
buildings in the south, which are of very recent date. ' It may, however, 
suffice to mention that in Ingalliko fjord itself, said to be experiencing 
a positive movement, high-level strand-lines are equally present, as in the 
north. 

The level of the sea depends on the tides, the heat of the sun, the 
atmospheric pressure, the prevailing winds, the influx of fresh water, and, 
in enclosed seas, on evaporation. It is also affected by local sources of 
attraction and many other circumstances. Among these, however, are 
some which make it difficult to ascertain the mean level, and render a long 



CH. I] OCEANIC SUBSIDENCES 27 

series of measuiements neoessaiy before changes of level can be made 
manifest; this is particularly true of the winds and climatic factors 
generally. Som^ movements, snch as the tides, may be easily eliminated, 
owing to their periodicity; others, again, such as the attraction of the 
eootinents, only need be considered in the case of particular telluric 
dianges ; others, finaUy, such as the deposition of sediments, act slowly 
bat constantly, producing a positive movement, which is general so far 
as it results from displacement, and local so far as it is a consequence 
of attraction. We shall have occasion later to study some of these factors 
in greater detail. 

In addition to all these processes great and general negative movements 
are from time to time produced by the formation of fresh oceanic abysses, 
or by the addition of new areas of subsidence to abysses already in 
existence, and it is important to bear in mind that movements of this kind 
surpass all others in importance. 

In the first volume of this work certain seas, in particular the Mediter- 
ranean, were studied in some detail ; the various ages of successive 
subsidences in the Aegean, the ^orth Atlantic, and the Tyrrhenian sea, 
were determined within narrow limits, and the relation of these seas to 
the older extensions of the Mediterranean were discussed. It is now my 
task to investigate the boundaries, first of the Pacific, and then of the 
Atlantic Ocean, and to compare them with each other, as a preliminary to 
the discussion of the great oceanic areas of subsidence. While thus 
oigaged in a study of the coasts, opportunities will arise of completing 
the tectonic sketches which were outlined in the first volume, and so of 
preparing the way for the survey of the face of the earth, reserved for the 
last volume. 

Although the formation of the Aegean inbreak falls possibly within 
a period more recent than the glacial episode, when the region was already 
inhabited by man, still we may fairly assert that a general subsidence of 
the strand-line, as a consequence of the sudden formation of a great oceanic 
abyss, has not taken place for some thousands of years past. Tet in 
contemplating those trifiing changes in the coast-line, which take place 
here and there before our eyes — ^the results, as we have seen, of a great 
variety of circumstances — ^we are again and again tempted to regard them 
as the direct successors of those remote events which have left their traces 
in a long-forgotten past. Thus the displacements of the strand-line in 
the Baltic, at present a closed sea, have been interpreted as the efiects of 
the same causes as gave rise to the ancient high-level strand-lines and 
shelly beaches of a sea which was certainly not thus dosed. 

It thus becomes necessary to subject some of the more important cases 
of change observed during the historic period to a searching examination. 
If, however, the result should lead us to the conclusion that local factors of 



28 DISPLACEMENT OF THE STRAND [part m 

a climatic nature, for instance, possess a greater influence than is usually 
attributed to them, and that where these are absent long stretches of coast 
show no signs of movement within the historic period, yet this would not 
prove that such movement does not take place. A study of the animfti 
kingdom does not reveal any change in species within the historic period^ 
but it does not follow that species are immutable. All that we can 
conclude is that within the limits of observation and of the period over 
which our observations extend, such changes are not discernible. But 
this result would involve considerable modification in current views sjs 
regards the strand-line. 

With the increased attention given during the last few years to the 
study of earthquakes and dislocations we have gained a closer acquaintance 
with the characters which distinguish a dislocation of the lithosphere, and 
at the same time have come to recognize the rarity of such an occurrence* 
The most remarkable dislocations of our times are no doubt those in the 
west of the United States, which have been described by Oilbert on the 
borders of the Great Salt lake, by Russell in the Oreat Basin, and by 
Reyer in the Sierra Nevada. These are all changes in the relative position 
of two segments of the earth's crust, occurring along a line which i$ 
usually many miles in length. The throw does not as a rule amount to 
more than. a few metres ; the course at the surface of the ground is always 
very sharply defined, often indeed it appears as an open cleft many miles 
long. 

I know, however, of only a single case in which a recent dislocation 
has reached the sea-coast and produced a local change in the strand-line. 
This occurred in New Zealand : Lyell has given an account of it \ 

In the year 1848, during an earthquake at White Bluff, Cloudy bay, 
a fissure opened on the south coast of Cook strait, and ran to the south- 
south-west, parallel to the course of the mountains; it was said to be 
traceable into the interior for a distance of 60 miles. On January 28, 
1855, a violent earthquake again occurred, and this time a line of 
dislocation appeared on the north coast of Cook strait; it began on the 
east side of Muka-Muka cliff, 12 miles south-east of Wellington, and 
was continued in a north-north-easterly direction, following closely the 
eastern foot of the Bemutaka chain ; it presented, for the greater part of 
its course, the form of an open fissure, and extended 90 miles into the 
interior of North island. This fracture may perhaps be regarded as 
the continuation of that which was formed in 1848 in South island. The 
area lying to the east of this line remained wholly unaffected by the fault 
of 1855 ; that to the west was displaced, downwards in South island, and 

' C. Lyell, Bull, soa g6oL de Fr., 1865, 3« B^r., XIII, pp. 661-667 ; Principles of Geology, 
nth ed., 1872, pp. 82-89 ; also R. Mallet, Report on the Facts and Theory of Earth- 
quake Phenomena^ Rep. Brit Ass., 1858, p. 105. 



CH. I] DISLOCATION IN NEW ZEALAND 29 

upwards in North island. In the south the subsidence close to the fissure 
amounted to 6 feet, in the north the elevation at its maximum, on the east 
side of Muka-Muka cliff, was 9 feet: a white band of nullipores which 
marked the former strand now stood, west of the line of dislocation, 9 feet 
above the water-level, while to the east its position remained unchanged* 

From here the movement decreased towards the west, so that at Port 
Nicholson, 11 miles to the west, it was reduced to one-half, and at a distance 
of 23 miles was no longer observed. In this case we may speak of a tilting 
movement, for the dislocation is visible not only on the coast but in the 
interior of the island, and the facts are beyond doubt. The presence of 
a visible line of dislocation sharply separating the moved area from the 
unmoved distinguishes this, case from all the others which we are about 
to discuss. 

The method we shall adopt in this discussion is as follows : — 

We shall first explain the structure of the shores of the Atlantic and 
of the Pacific Oceans, and then point out the very remarkable contrast 
which distinguishes these two ocean-b&sins. 

We shall next pass on to consider the seas of ancient times. It must 
not be forgotten that it was the extraordinary extent of the seas in the 
eariier periods of the earth's history which gave the first impulse to aU 
these discussions. The question yrill then arise whether the nature and 
the distribution of the ancient sediments point to local or general altera- 
tions in the sea-leveL For the solution of this it will not be necessary to 
treat the deposits of each stratified system with equal fullness. On the 
contrary, I do not propose to discuss the distribution of the sediments in 
any detail except in those cases where observations have already led to 
clear and definite results, and I shall only incidentally refer 'to their nature. 
Thus in the Carboniferous the presence of the Coal-measures, so far as they 
bear on the question before us, wiU be taken into consideration, and in the 
Rhaetic period the formation of the limestones. 

Among the events of the immediate past and of the existing present 
are some of exceptional importance, such as the origin of the Norwegian 
strand-lines, the behaviour of the Baltic, and the temple of Serapis at 
Pozzuoli in the Mediterranean ; these demand a more detailed investigation. 
In oondnsion we shall attempt a survey of existing observations. 



30 



CHAPTER II 
THE OUTLINES OF THE ATLANTIC OCEAN 

The Canadian shield. The Baltic shield. Glint lines. The table-land of Spitzbergen. 
Greenland. The Caledonian mountains. The Armorican mountains. The Tariscian 
mountains. The syntazis of Central Europe. The Iberian Meseta. Surrey of the pre- 
Permian mountains in Europe. The islands of Europe. Western Africa. The east of 
Central and South America. Survey of the outlines of the Atlantic. 

The Canadiaii Shidd. The whole of the north-east of America, from 
the mouth of the' St Lawrence to that of the Mackenzie, together with 
the adjacent islands of the Arctic Ocean, belongs to a broad table-land of 
hori^ntal Palaeozoic bedsj from beneath which the Archaean foundation 
crops out in the middle of the table-land not unlike a flat shield. This 
Archaean shield is thus surrounded by a ring of horizontally stratified sedi- 
ments. The primaeval rocks composing it were not only folded in pre- 
Cambrian times, but were also exposed to severe denudation, so that the 
Palaeozoic series rests on the planed-down edges of Archaean folds. Many 
great patches of the Palaeozoic covering, however, remain preserved on the 
shield itself. The exposure of the shield, the outlines of the inner maigin 
of its Palaeozoic girdle, as well as of the superincumbent Palaeozoic patches, 
result to a great extent from the glacial erosion which these regions have 
experienced in comparatively recent times. 

It is to the exposed Archaean surface that we give the name of the 
Canadian shield. 

Resting upon this shield, a little outside the central region towards the 
east, lies the sheet of water known as Hvdson bay. As we might expect 
from the uniform structure of the land, this large arm of the sea is of a very 
uniform and trifling depth, which amounts, in Hudson bay and James 
bay, to only about 70 fathoms ; towards Hudson strait, the bottom sinks 
to 100 fathoms and over ; in Fox channel the depths are much more con- 
siderable. Hudson bay cannot, therefore, be compared with the deep 
inbreaks of the Mediterranean, the gulf of Mexico, or the Caribbean sea. 
Neither the expression ' fore-sea ' (Vormeer) nor ' back-sea ' (BUckmeer) is 
here applicable ; the bay is only a submerged plateau, a shallow pan. 

A similar structure is presented elsewhere in only one instance, that of 
the Baltic sea. 

Such details as we possess concerning the structure of the neighbourhood 
of Hudson bay, and of the several Palaeozoic patches which rest upon the 



CH. II] THE CANADIAN SHIELD 31 

shield, we owe almost exclusively to the labours of Robert BelP. The 
shores are flat ; only on the east coast of James bay and to the north of it, 
towards cape Wolstenholme, do more considerable heights occur, attaining 
nearly 2,000 feet 

*A band of horizontally stratified sediments of Silurian and Devonian 
age starts from the south and extends for a great distance up to the southern 
and western part of the shores of James bay. On the west of Hudson bay, 
near the month of the Churchill river, lies a great mass of ancient quartzites 
with horizontal stratification. In the north crystalline schists are found 
about Chesterfield inlet ; Marble island owes its name to an e^^r ; it is not 
formed of marble, but of quartzite and schist. At cape Southampton and 
on Mansfield island thinly stratified fossiliferous limestone of Silurian age 
lies in hundreds of horizontal beds, which the undermining waves break up 
into towers and pillars. The contrast between the plateau-like Mansfield 
island and the surrounding country is described as striking in the extreme. 
Between these regions the greater part of the land consists in all proba- 
bility of gneiss. Cape Wolstenholme is also formed of gneiss, as well as 
a large part of the east coast, but in Natapuka sound remains of the 
Palaeozoic covering occur. 

We will now pass out of the bay through Hudson strait, where the 
border of the shield is least known and where its form appears to be least 
regnlar. 

The islands off cape Wolstenholme, Digges, Nottingham, and Charles 
island, as well as Prince of Wales sound and North bay lying opposite, 
on the coast of Meta incognita, are wholly composed of gneiss, as is all 
that is known of the south coast of Meta incognita as far as the Savage 
islands, near Resolution island. 

Proceeding towards cape Chudley and the sea, however, we encounter 
an independent and fairly lofty chain, which can no longer be regarded as 
belonging to the Canadian shield or to its border. The position of this, as 
we shall see later, is of particular importance : it begins near Belle Isle straits 
and forms the whole coast of Labrador; in its northern part it attains 
a height of 6,000 feet and reaches Hudson strait near cape Chudley with 
a height of 1,500 feci According to Bell's observations this coastal chain 
is not broad, and is followed on the west by a fiat country through which 
the Whale river flows to the north into Ungava bay. In the south the two 
Kttle islands, Castle island and Henley island, facing the northern point of 

^ & BeU, Report on Hudion'i Bay and lome of the Lakes and Riven lying to the West 
of it» Geol. Sorv. Canada, Report for 1878-1880, p. 27, C. et aeq. ; by the same. Report on 
the Geology, Mineralogy, Zoology, and Botany of the Coast of Labrador, Hudson's Straits, 
and Hndaon's Bay, op. cit. for 1882-1884, DD. Digges, near cape Wolstenholme, is 
only a single island cat through by two deep farrows ; it was at one time supposed that 
three independent islands existed here. 



32 THE OUTLINES OP THE ATLANTIC OCEAN [pabt ni 

Newfoundland, consist of horizontal sheets of an eruptive rock such as are 
known in several places elsewhere in Canada, intercalated into the lowest 
Palaeozoic beds, and these sheets form a startling contrast to the surrounding* 
country, which consists of gneiss. We may add at once that the island of 
Anticosti in the gulf of St. Lawrence is formed of flat-lying Silurian beds. 

The coast-chain of Labrador consists, so far as it is known, of gneiss and 
other Archaean rocks. Its lower parts are rounded by ice, but above it is 
broken up into jagged peaks and sharp ridges. The ice did not reach the 
upper parts, although in the south it extended up the slopes to a height of 
about 1,600 feet. 

Sutherland had already shown that the highly glaciated west coast of 
Davis strait and Baffin bay, from Cumberland bay nearly as far as cape 
Walter Bathurst, consists chiefly of Archaean rocks ; traces of a Palaeozoic 
covering have only been met with near cape Durban and to the south of it ^. 
Hall subsequently visited the southern part of these regions: the most 
detailed observations have been made by Boas. From these it appears that a 
high and narrow coast-chain, precisely similar to that of Labrador, exists here. 
Dr. Boas has had the kindness to furnish me with the following description : 
' The narrow range which forms Cumberland peninsula runs along the west 
coast of Baffin bay up to Lancaster sound. In Home bay, where the coast 
assumes a north to south direction, a remarkable gap occurs, and the hilly 
country of the west advances as far as Davis strait. This range, as far as 
it is known to me, consists in its central part of gneiss, in the peripheral 
parts of coarse-grained granites. The whole range, rising in steep horns 
and peaks to a height of 2,000 meters and over, is characterized by narrow 
valleys with precipitous walls which extend across the peninsula and con- 
nect the corresponding fjords of the opposite coasts, the height of the passes 
being scarcely 160 meters. The country is thus cut up into steep-walled 
mountain masses, three of which succeed each other as far as Home bay. 
Further to the north also the mountains are completely intersected by deep 
fjords which merge into vallejns opening on to the western plains. Pre- 
cisely analogous features are met with in the extreme north, where Hayes 
sound appears to form a similar divisional line, and where the valley which 
connects Greeley and Archer fjords separates two mighty highlands. 

' To the west of the range lies a hilly country, apparently quite irregular, 
and composed of coarse-grained granite. A glance at the fjords of the 
north-west coast of Cumberland shows the predominance of a north-west 
and north to south direction in the course of the valleys. The form of the 
Qords is here very striking: they consist almost everywhere of basins 
connected by narrow gullies in which rapids are formed by the water 
flowing into or out of them according to the state of the tide. 

^ P. G. Sutherland, On the Geological and Glacial Phenomena of the Coasts of Davia 
Strait and Baffin's Bay ; Quart. Joum. GeoL Soc.. 1853, IX, p. 299. 



CH. n] CUMBERLAND IN BAPPTN-LAND 83 

* As we advance towards the west the hills become lower and the valleys 
broaden until we find ourselves at last on a boundless plain. 

' Here the Silurian limestones b^^in, which in the more southerly parts 
of the country had been found at the upper end of Frobisher bay and on 
the shores of lake Kennedy, where they are surprisingly rich in fossils. 
The first of these localities was made known by Hall. Unfortunately 
I did not find the rock in place, since the plains were covered by very deep 
snow, so that I cannot say whether the beds lie horizontal or not; 
probably they do. The lakes of this r^on must certainly be regarded 
as relics of the sea. The whole eastern half of Fox basin is flat, and 
the plain is an exposed sea bottom, as is proved by the remains of whales, 
walruses, &c. 

' Oiographically, the peninsula bounded by Cumberland sound and 
Frobisher bay is completely separated from the range described above. 
The peninsula attains its greatest height in its southern parts, and slopes 
away to the plain towards the north-west. The northern shore consists 
exclusively of granite: in the south limestones (Silurian?) occur. Sand- 
stone also has been found in the extreme south. 

* The peninsula of Meta incognita is also completely independent of the 
northern plateau, for the plain reaches as far as Frobisher bay.' 

I have given this information in full, since it embodies all that we 
actually know of a lofty and independent range ; consisting chiefly of gneiss 
and extending from the southern part of Cumberland nearly as far north 
as cape Walter Bathurst. It is either the direct continuation of the lofty 
gneiss range, which Robert Bell followed along the coast of Labrador from 
Belle Isle strait to cape Chudleigh, or at least corresponds very closely in 
position with such a continuation. 

To the west of this coastal range lies a flat country, in which Silurian 
limestone with horizontal stratification has been observed in several locali- 
ties. Boas conjectures, with great probability, that the horizontal Silurian 
limestone which we shall meet with later more to the north in Prince Regent 
inlet is continued thence on to the flat eastern half of Melville peninsula, 
where Hall observed it near Igluling, thence to lake Nettilling (lake 
Kennedy) and to the upper end of Frobisher bay K Such then is the north- 
eastern border of the Canadian shield, separated by a long and lofty gneiss 
range from the depths of Baffin bay and Davis strait. 

A continuation of this Silurian zone is so far not known on the shores 
of Hudson strait, and it is therefore not possible to say whether it should 
be looked for in the plain which extends from Unga va bay through Labrador 
southwards to Belle Isle strait and the island of AnticostL 

That Anticosti consists of horizontal Silurian limestone has already been 

* Fnns Boas, Baffin-Land ; Geographische Eigebniase einer 1888 and 1884 aaagefilllirten 
FonchangsreiBe: Peteim. MittheiL, Eigftnzungsheft No. 80, 1885, pp. 50 and 57. 
lOLiiM n n 



34 THE OUTLINES OP THE ATLANTIC OCEAN [part in 

mentioned ; but just as to the north-east the Silurian girdle in Cumberland 
peninsula is bounded on the exterior by the lofty gneiss range of the ooasb, 
so on the St. Lawrence it is bounded on the exterior by the folded mass 
of Maine and New Brunswick. 

The east of the United States is traversed by folds which have been 
produced by a tangential movement directed from the existing Atlantic 
ocean towards the . mainland. We have seen (L pp. 558-555) that from 
Alabama to Oeorgia these folds strike to the north-east more or less 
parallel with the coast, thus forming the Alleghany mountains, and are 
finally continued into the Catskill mountains north-north-west of New 
York, where they are joined by a system of folds striking from north to 
south. 

The Alleghanies are not separated from the adjacent plain on the west ; 
the intensity of the folding decreases in this direction and a secondary 
folded formation or parma, the Cincinnati uplift, lies in front of them. 
With the folds which run from north to south the case is different Their 
western border extends along the length of lake Champlain, reaches the 
St Lawrence near Quebec, and then, with the trend curving to the north- 
east, follows the right shore of this great river. The country to the right 
of the river is folded, that to the left of it is flat table-land. 

It is true that on the left shore the whole Archaean and Azoic series 
has been intensely folded, but it was abraded at a very early period, and <m 
the denuded folds several members of the Silurian were deposited horizon- 
tally ; these, however, have been so far removed in later times, that over 
large areas the Archaean foundation alone is visible. This is the case 
according to Laflamme around lake St John and in the whole fluvial 
region of the Saguenay. Some parts of the Palaeozoic patches of lake 
St John contain petroleum ^. . 

* 

Selwyn's admirable account of the structure of this region confirms 
Logan's view, that the St Lawrence must be regarded as marking an 
extremely important boundary line in the structure of the country \ All 
the Silurian on the right bank, from Quebec away towards cape Rozier, 
follows the trend of the shore-line and is overfolded towards the west, so 
that the beds are inverted as in the outer border of the Alpa These Silu- 
rian folds are followed towards the east, at a distance of a few miles from 
the river, by a long and narrow but not continuous zone of Archaean rocks ; 
two other similar zones, running close together and nearly parallel, strike 
north and south towards the south-east part of Chaleur bay ; another zone 

^ J. C. E. Laflamme, Report on (xeological Observations in the Region of the Saguenay ; 
Oeol. Snrv. Canada, Report for 1882-1884, D. 

* A. Selwyn and G. M. DawBon, Bescriptive Sketch of the Physical Geography and 
Geology of the Dominion of Canada, 8vo, Montreal, 1884, with map, pp. 5-26; alao 
E. Gilpin, jnn., The Geology of Cape Breton Island, Nova Scotia, Quart. Joum. Geol. Soe^ 
1886, XLII, pp. 515-526, map. 



OH.n] THE LOWEE ST. LAWRENCE 35 

o{ Archaean masses nms in a curved line along ih.e bay of Fondy, then 
readwe the most northerly port of cape Breton on the north-east and is 
oontsQoed across the sea into Newfoundland. Still farther to the east 
follow the zones directed to the east-north-east which have been already 
mentioned (L p. 554). 

The folding is older than a part of the Carboniferous period, as is 
shown by the lie of the Coal-measures in New Brunswick, and did not 
extend beyond the present line of the lower St lAwrence, but it has 
been dammed back against this line, and overfolding has taken place in 
conaeqaenca Selwyn represents the structure of the St. Lawrence valley 
itself by the aooompanying section. 

On the left we see the edge of the table-land ; an upper member of the 
lower SUnrian (Trenttm limestone) lies in tran^reesion on the abraded 



SL lAwruiM 



FiB- 8. HgpollutKal Stctfon ocrotB the ritvr St. Laienmee and Hit iah (/ Orteam. After Seliryn. 
I, II, III Faults ; gn Gnaisi ; / red, green, uid bliok sliAleB and nndatones with Cunbriut 
foBaila In the incladed pebbles (' Lauzon Btue ') ; 3 Lerie oonglomerata and Qraptolite atulea 
(Lower SUurianl ; j Chazj limestone and other beds (Lower Silurian) ; 4 Trenton limestone, 
tjring tranagreiaiTelj on the gneiM (Upper Silurian) ; j, 6 Utica and Hndsoa stagea (Upper 
Silurian]. 

folds of the gneiss. The northern arm of the St. lAwrence is & true 
&alt trough, for the downthrown sbrata of the river bed, visible along the 
fractures on both shores, belong- to the more recent divisions of the lower 
Silurian (Utica and Hudson stages) ; the isle of Orleans is a horst, and at 
the same time part of the folded region. 

The boundary between the folded region and the table-land runs 
bc^een cape Bozier and the island of Anticosti ; this part of the gulf of 
St Lawrence and the lower part of the river is therefore a true 'fore- 
valley ' (Vorthal), like the Fetsian gulf. 

The folds of New Brunswick and cape Breton island run out in ja^ed 
ends towards the gulf of St. Lawrence, and are continued, as we said 
above, into Newfoundland. The very first examination of the island made 
by Jokes in the years 1839 and 1840, showed that its outline is entirely deter- 
mined by folds striking across it; the saddles of these folds form peninsulas 



< 



36 THE OUTLINES OP THE ATLANTIC OCEAN [pakt m 

both on the north and south coasts, while the bays, such as White bay, Notre- 
Dame, Bonavista, Trinity and Conception bay in the north, and Fortune, 
Placentia and Si Mary bay in the south, correspond, for the greater part, 
to synclinals. The later surveys of Murray and Howley confirm this result ; 
the great northern peninsula is probably a single continuous anticlinal ; 
a fault appears to run with the strike from the neighbourhood of cape Bay 
in the south-west of the island through the Grand lake to White bay, that 
is, across the whole island : in the south-east, the extent to which the out- 
line is determined by the folds is shown with particular clearness by the 
Avalon peninsula. The mean direction of the folds is N. %T E. They are 
formed of Archaean and Palaeozoic rocks; the Carboniferous formation, 
or at least a part of it, lies unconf ormably upon them ^. 

Here we have before us a striking example of that form of coast which,, 
with F. von l^ichthof en, we will term the mea coast. It indicates the plung- 
ing of great folded mountains beneath the ocean \ Thus, the outer border 
of the girdle of flat-lying beds surrounding the Canadian shield is defined 
towards the north-east by the gneiss mountains of the west coast, by 
Baffin bay, Davis strait, and Labrador; towards the south-east by the 
St. Lawrence, and the folded mountains of New Brunswick and New- 
foundland. On the south and west there is no such outer boundary of the 
encircling zone ; it becomes continuous, in Michigan, Wisconsin, Minnesota 
and the northern Prairie land, with the widely extended Palaeozoic deposita 
of the neighbouring regions, and is covered towards the west by the 
transgressive Cretaceous. 

On the inner border of the girdle, that is, along the margin of the 
Archaean shield, a phenomenon appears for the first time to which I shall 
subsequently have firequent occasion to refer ; I €illude to the evident con* 
nexion which exists between the position of the great lakes and the course 
of this boundary line. 

The line runs from the north-east end of lake Ontario to the east end 
of Georgian bay, and thence towards the west end of lake Superior. The 
northern shores of the great sheet of water consist chiefly of Archaean, 
or at least Azoic rocks, while on the south side the various members of the 
Palaeozoic series occur, and spread out towards the United States. 

Further to the north-west, the coincidence of the lakes with the boundary 
of the Archaean shield is still more striking ; Richardson, Isbister, and other 
observers long ago called attention to'it^. 

^ J, B. Jukes, General Report of the Geological Survey of Newfoundland, 8vo, London, 
1848, map, p. 128 et seq. ; A. Murray and J. P. Howley, Geological Survey of Newfound- 
land, 8vo, London, 1881, p. 189 et passim ; also J. Milne, Notes on the Physical Nature 
and Mineralogy of Newfoundland, Quart. Joum. GeoL Soc^ 1874, XXX, pp. 722-745. 

' F. Ton Richthofen, ^"Qhrer fdr Forschungsreisende, 8vo, 1886, p. 308. 

' J. Richardson, On some Points of the Physical Geography of North America in 
connexion with its Geological Structure, Quart. Joum. Geol. Soc, 1851, VII, pp. 212-215; 



<H.n] CANADIAN ZONE OF LAKES 37 

• 

The east side of lake Winnipeg is Archaean, while the island and the 
west shore belong to the horizontally stratified Palaeozoic border; on this 
border lie Winnipegosis, Manitoba, and a crowd of smaller sheets of water, 
4knd to the west of these begins that vast region of Cretaceous deposits 
which extends to the foot of the Rocky mountains (I, pp. 558, 584). 

In descending the river Athabasca, Bell found himself between 
Cretaceous rocks as far as Drowned rapid, ie. to about lat 66** 4ff N. ; 
firom this point onwards Devonian sandstone, rich in petroleum and lying 
horizontal, makes its appearance. The petroleum forces its way up through 
the Cretaceous beds, flows out and covers the slopes of the shores with 
bitumen. The outflow, however, appears to be very local ; a thick bed of 
•day at the base of the Cretaceous holds the petroleum down over large 
areas. Some patches of Cretaceous are still to be seen further to the 
north ; the flat-lying Palaeozoic deposits are continued to lake Athabasca 
and form its southern shore as far as it is known. The northern shore, 
on the other hand, consists of gneiss. Gneiss also forms the islands at the 
west end of the lake and the country about the mouth of the river 
Athabasca, and the exit of the Great Slave river \ 

The work of the Geological Survey. of Canada, which has extended as 
far as the Great Slave lake, and the highly important explorations of 
Richardson, Toule Hind, Eennioott, and others, in the inhospitable regions 
which stretch northwards from the Great Slave lake to the Arctic Ocean, 
enable us to trace the boundary between the Archaean shield and the 
Palaeozoic border, which takes approximately the following course * : — 

For a very long distance it almost coincides with the Slave river ; and 
that half of the Great Slave lake which lies east of the mouth of this river is 
floored by Archaean rocks, that to the west, on the other hand, by Palaeozoic 
beds. On the west shore important springs of petroleum occur. To the 
north of the Great Slave lake the boundary is indicated, according to 
Richardson, first by a long arm of the lake, then by a series of sm€iller 
sheets of water. It extends to lake La Martre, then with a change of 
direction reaches the south-east part of Great Bear lake in MacTavish 
bay, and foUows the lower course of the Coppermine river down to the sea, 

1>7 the same, NarratiTe of an Arctic Search Expedition, 2 yols., London, 1851, and The 
Polar Begions, Edinburgh, 1861, pp. 285-289; Murchison, Siluria, 4th ed., 1867, p. 440 
et paasim ; further, k. K. Isbister, On the Geology of the Hudson Bay Temtories and 
of portions of the Arctic and North-western Regions of America, Quart. Joum. Geol. Soc, 
1855, XI, p. 503. This western boundaiy of the shield from the south up to the Slave 
lalce is marked on the general map of the Greological Survey of Canada which appeared 
in 1883. 

> B. Bell, Report on Part of the Basin of the Athabasca River, N.W. Territory ; Geol. 
Snrv. Canada, Report for 1882>1884, CC. 

» F. B. Meek, Remarks on the Geology of the Valley of Mackenrie River, with Figures 
and Desciiptions of Fossils from that Region, chiefly collected by the late R. Eennicott; 
Tnuis. Chic Acad. Sci, 1867-1869, 1, pp. 61-114, pi. 



38 THE OUTLINES OF THE ATLANTIC OCEAN [PAUTnr 

80 that cape Barrow and its adjacent islands in Coronation golf consist of 
granite and gneiss, while cape Krosenstem, a little farther west, belongs* 
to the Palaeozoic girdle, which, maintaining a great breadth, extends beyond 
the mouth of the Mackenzie. 

Those explorers who have followed the course of the Mackenzie have 
thus encountered only Palaeozoic deposits and some patches of upper Cre- 
taceous and Tertiaiy ; but we have already mentioned that from the upper* 
tributaries of the Peace river the border of the Rocky mountains assumes 
an almost north-and-south direction, so that it should meet the Mackenzie 
below the confluence of the liard (I, p. 558), and indeed, steeply upturned 
beds, offering a striking contrast to the horizontal strata of the rest of the 
broad border, have been observed at several localities within this region, 
particularly below Fort Simpson, that is to say a little below the con- 
fluence of the liard and the Mackenzie, and the heights visible from there 
have^ in fact, been expressly described as spurs of the Rocky mountains. 
Then follows on the Mackenzie river a long stretch of country formed of 
horizontal Palaeozoic beds; near the Ramparts, however, in lat. 66"" N. 
a steep dip to the north-west predominates, but still further down the 
valley, in the narrows above the delta, the beds again become horizontal ; 
to the east they are known as far as the upper course of Anderson river,, 
or indeed, as we saw above, even as far as cape Ejrusenstem, and to the 
west as far as the upper reaches of Porcupine river. Eennioott, MacFarlane,, 
and Petitot have traced them in these remote regions ^. 

Some connexion with the Rocky mountains thus evidently exists, though 
its precise nature is not known. A very remarkable fact, established by 
Meek, is the extraordinaiy constancy and importance presented by the 
middle divisions of the Devonian within the Palaeozoic border. The f ossila 
which have been found in it enable us in fact to trace the deposits of the 
Hamilton group (middle Devonian) from Rock island, Illinois, through 
Iowa, Minnesota, Dakota, the great girdle along the lakes, and down the 
Mackenzie almost to the Arctic Ocean; the fossiliferous localities* most 
remote from one another, separated by nearly thirty degrees of latitude, 
possess a not inconsiderable number of species in common ; in the souths 
as well as at several localities in the extreme north, these deposits are 
characterized by their richness in petroleum and by salt springs; this, 
character extends, indeed, as far east as Oasp^, and we have already men- 
tioned it in connexion with the isolated patches of lake St. John, north 
of the lower St. Lawrence. 

As regards more recent deposits a number of patches may be observed 
on the Bear river, which flows from the Great Bear lake into the Mackenzie. 

^ Meek, torn, cit., p. 74 et seq. ; Hubert, Bocuments sur la g^logie du MacKenzier 
recneillis par le P^re Petitot, Bull. Soc. g^ol. de Fr., 1874-1875, 8« s^r.. Ill, p. 87 ;. 
Petitot, Notes gtelogiques sur le bassin da MacEenzie, torn, cit, pp. 88-93. 



ck. n] CANADIAN ZONE OF LAKES 39 

Sir Jdtm BichordBon found an Ammonite in the rapids of this river; Hind 
brought home an Ammonite and an Inoceramns, obtained from horizontal 
beds. The marine Cretaceous beds thus extend along the outer border of 
the Boeky moantains from the far-distant south to lat. 66"* N. Tertiary 
Ignites appear in many places, both on the Bear river and the Maekenzie; 
the plant remains have been described by Heer. They extend as far as 
the islands which lie opposite the mouth of the Mackenzie ; one patdi lies 
between cape Bathurst and cape Parry ; it was this patch that Mierteching 
saw on fire in 1850. From here the lignites extend further to the west 
ooast of Banks' Land and Prince Patrick island \ The phonolite mountains 
of the lower Mackenzie and of the Bear river, mentioned by Petitot, are 
probably of the same age. 

We thus perceive that the abraded Archaean platform, on which 
Hudson bay lies, may be traced from the east to a great girdle of lakes, 
the position of which stands in evident connexion with the course of the 
flatly-bedded Palaeozoic border. For, apart from the isolated patches of 
Palaeozoic resting upon it, already mentioned as occurring on the shores of 
Hudson bay, the boundary of the Archaean region runs in such a manner 
that the north coast of the northern bays of lake Huron, the north coast 
of lake Superior, and the east coast of lake Winnipeg still consist of 
Archaean rocks, while the sheets of water themselves lie entirely or for 
the greater part on the Palaeozoic girdle; and from lake Winnipeg the 
boundaiy runs through the western end of lake Athabasca across the Great 
Slave lake, following the great notch in its north side near Fort Bea, 
through lake La Martre, through the east half of the Great Bear lake, and 
finaUy to Coronation gulf in the Arctic Ocean. 

Isbister long ago pointed out the analogy between the position of 
CorcmatiQn gulf and of the lakes just discussed. In order to examine 
more closely into this resemblance, let us now turn to the Arctic 
archipelago. 

Among the many perilous journeys made through the Arctic archi- 
pelago of North America, none has produced richer results, as regards 
our knowledge of the geological structure of this region, than the search 
for the remains of Sir John Franklin and his companions, conducted by 
M'Clintock in the years 1857 to 1859. It was the results of this expe- 
ditkm that Haughton made use of, in conjunction with other observations 
of an earlier date, when he constructed the first general map of the region, 
and clearly displayed the extremely simple manner in which its stratified 
rocks are arranged ^ 

^ O. Heer, Flora fossilia arctica, 1868, 1, pp. 25 and 13.5-139. 

* Gapt F. L. M'Chiitock, Beminiscences of Arctic Ice Travel in Search of Sir J. Fiaaklin 
aid his Companions, with geological Notei and Illustrations by S. Haughton, Journ. R07. 
Dub. Soc^ 1858, J, pp. 183-250, pL and geological map ; by the same, The Voyage of the 



40 THE OUTLINES OP THE ATLA2«TIC OCEAN [pake in 

All these islands and peninsnlae, ^m ihe continent of America to the 
north of Parry islands, form the horizontally stratified ncorthem border of 
the great Archaean shield which we have jost studied in the north of the 
continent : the several sedimentary formations of tills border are so arraiiged 
that they strike to the east or north-east, and become progressively more 
recent as they are traced towards the pole *, 

The Archaean rocks, which form cape Barrow and the adjacent ialaoda 



Fio. 4. Arctic SoTlh Amiriea (ftfter Hsclure, Haughton, uid othen). 
A Archaean and granite ; ( Silnrian ; at Devonian (?) and baial Cart>oniferouB ; ea Car- 
boniferoQB limMtone; m Trias (?) and Juraaiic ', tg Leaf-bearing Tertiary beds (cape BathuTSt, 
north-west Banka' Land and Prince Patrick island); Basalt on cape Alexander in Smith sound. 

in Coronation gulf, are also met with at the moutii of the Great Fish 
river; Bae observed them on Melville peninsula, and their northern mfurgin 
passing through EcUpse sound reaches the shore of Baffin bay near cape 

Fox in the Arctic Setu— A Horrative of the Discorerj of the Fate of Sir J. Franklin 
and hii Companions, Svo, 1859 ; Append. No. IV, Geolo^cal Account of the Arctic 
Archipelago, bj S. Haughton, pp. 372-399, and geological map; general desoriptiona 
also in 0. Heer, Flora fossilis arctica, I, 4to, 1868, map ; C. E. de Ranee, Arctic Geology, 
Nature, 1875, p. 448, geological sketch-map. The most important works are cited by 
R. Jones, Manual of the Natural History, Qeology and Physics of Greenland and the 
neighbouring regions, 6to, 1875, London. 

' A new map of the Arctic regions by G. U. Dawson appeared in 1883, too late to be 
made nse of here. It does not render necessary any important alteration of oar genend 
description, bnt extends the Silurian, for example, over WoUaston lAnd and Victoria 
Land ; G. M. Dawson, Notes to accompany a Geological Map of the Northern Portion of 
the Dominion of Canada, Oeol. Suiv. Canada, 1887 (Annual Report for 1886), S, 62 pp. 
and map. 



€H. n] NORTH EDGE OP CANADIAN SHIELD 41 

Walter BathniBt To the north of this cape, the eastern part of North 
Devon as far as cape Warrender consists of the same rocks. 

In these regions, especially near cape Walter Bathurst, as well as in 
Wolstenhokne sound on the east side of Baffin bay, the ancient rocks are 
overlaid by red sandstone, which is, perhaps, the same sandstone as that 
visible towards the west, in North Somerset, as the base or lowest member 
of the Silurian deposits. 

The Airhaean region is followed by a broad girdle of Silurian, which, 
in continuation of the broad Palaeozoic region already mentioned as lying 
between cape Erusenstem and the mouth of the Mackenzie, extends from 
Ck>ronation golf to Baffin bay. This zone consists of alternating beds of 
limestone and argillaceous sediments: their horizontal stratification gives 
to the cliffi that resemblance to fortifications so often mentioned by Arctic 
explorers. 

With regard to WoUaston Land and Victoria Land I possess no informa- 
tion. The Silurian zone is known in the southern pa^ of Banks' Land, 
and it forms the northern half of Prince Albert Land ; it includes Prince 
of Wales Land, the little islands in Barrow strait, the greater part of 
Comwallis island and of North Devon up to Jones sound, then both shores 
of Lancaster sound and of Prince Regent inlet, North Somerset, and to the 
south the greater part of Boothia with the magnetic pole, and King William 
Land^ 

In Peel sound the Archaean or more recent granitic rocks are visible on 
the coast beneath the Silurian : cape M'Clure on Prince of Wales Land is 
described as eruptive syenite. 

It is very remarkable, considering the great development of Devonian 
sediments in the valley of the Mackenzie, that only isolated traces of 
a Devonian zone have been met with in the Arctic archipelago, as, for 
instance, on the Princess Royal islands and Byam Martin. 

The next zone is formed mainly of sandstone, and in some localities 
contains coal beds. Heer, judging by the position of the beds and the 
plant remains, assigns it to the lower Carboniferous and the Ursa stage ^. 
It forms the north half of Banks' Land, the south half of Eglinton, Melville 
island with the exception of the two northern peninsulas, Byam Martin, 
next Bathurst, with th^ exception of the three northern peninsulas, and is 
then, perhaps, continued across North Devon to the small islands which lie 
to the east. 

This ooal-bearing zone is followed in normal superposition by a richly 
fossiliferous zone of marine Carboniferous limestone. It comprises Prince 

' The most important source of information previous to M'Clintock on the Silniian 
faxmsk of this zone is J. W. Salter, On Arctic Silurian Fossils ; Quart. Joum. Qeoh Soc, 
1853, IX, pp. 312-817, pL 

* Heer, Flora fossilis axctica, I, p. 19, et passim. 



42 THE OUTLINES OF THE ATLANTIC OCEAN [pa«phi 

Patrick island, the northern part of Eglinton, all the northern promontories 
of Melville and Bathurst, the little islands of Penny strait and Grinnell 
Land'. 

The Carboniferous limestone is the northernmost zoob ; but in the easfe 
of Prince Patrick island, on the north-west coast of Bathurst and on 
several little islands north of Grinnell Land, small patches of later 
deposits rest upon it, and indicate a Mesozoic zone lying beneath the 
Arctic Ocean. The fossils consist of the remains of Saurians, Ammonites^ 
and a few bivalves. Neumayr has compared the remains of Ammonites 
from Wilkie point. Prince Patrick island (lat. 76^ 20' N.), with species of 
the middle Jurassic'. 

Since the parallel zones just enumerated strike east-north-east or 
north-east, the question arises whether they do not reappear on the 
western shores of Smith sound or of Kennedy channel. The structure 
of these coasts, as represented by Feilden and de Bance, certainly shows 
that here also Archaean, Silurian, and Carboniferous rocks are present, and 
that here too a north-easterly strike predominates, but the beds are 
upturned and folded, and we thus enter a region of fundamentally different 
structure \ 

We will only cast a hasty glance over these regions. 

The Archaean rocks of Ponds inlet, cape Walter Bathurst, and eastern 
North Devon form towards the north- the whole of the lofty coast of 
Ellesmere Land, and near cape Isabella they cross over to the east side 
of Smith channel, where they are covered by leaf-bearing Tertiary beds in 
the vicinity of port Foulke. 

Silurian deposits reach the coast with a north-east strike, and occupy 
almost the whole area lying between lat. 79"" and 80'' N. On Bache island 
they lie fairly flat on the syenite and granite foundation, but in Norman 
Lockyer island in lat. 79"" 25' N. the dip becomes steeper, and further on an 
anticlinal is formed. This Silurian zone, which terminatee on the west 
side of Smith sound in Scoresby bay, is continued with the same strike 
obliquely across this arm of the sect, and on its east side forms in all 
probability a great part of the floor of the Humboldt glacier and of 
Washington Land. It also embraces Petermann fjord, Polaris bay, and 
the interior of Hall's Land as far as Newman bay. 

North of the Silurian, region follow strongly folded ancient rocks, mica 

^ The Grinnell Land of certain travellers, separated from North Devon only by Arthur 
straits, and not to be confounded with Grinnell Land in lat. 80"* N. on the Kennedy 
channel. 

* M. Neumayr, Die geographische Verbreitnng der Juraformation ; Denkschr. k. Akad. 
Wistenach. Wien, 1885, L, pp. d4 and 141. 

* Gapt. H. W. Feilden and C. E. de Ranee, Geology of the Coasts of the Arctic Lands 
visited by the late British Expedition under Capt. Sir G. Nares; Palaeontology by 
R. Etheridge ; Quart, joum. Geol. Soc., 1878, XXXIV, pp. 556-689, map and pi. 



CH.n] THE CANADIAN SHIELD 4S 

adiistB and quartzite, known as the ' Bawson beds/ They strike north- 
north-east and tann. the west shore of Seoresby bay up to cape Oreswell in 
latw 82^40'N. On the other side of Bobeson channel they form all the 
land north of Polaris bay and Newman bay. 

Far to the north, beyond the Rawson beds, some Deyonian appears in 
Dana bay; on the north side of Grinnell Land, on Feilden and Parry 
peninsnla up to Clements Harkham's inlet, Carboniferous limestone is 
met with, precisely in the line of strike of the Carboniferous limestone of 
tiie Patty islands. Beneath the Carboniferous limestone and beyond it, up 
tq the farthest point hitherto reached, stretch the rocks of the Bawson 
beds. — 

The results of the preceding observations may be summed up in the 
following sketch. 

The Canadian shield, that abraded Archaean surface which he$m the 
shallow waters of Hudson bay, is surrounded by a ring of flat-lying 
Palaeozoic beds; in a few places this ring has been entirely destroyed by 
denudation, as in the south-east towards the lower course of the Saint 
Lawrence, but elsewhere it is continuous and often of great breadth. On 
the west the Devonian contributes to its formation, though the Silurian 
has not yet been observed at its base. The margin of the shield on the 
soath and west is accompanied by a long series of great lakes. Lake 
Huron with its northern bays, lakes Superior, Winnipeg, Athabasca, the 
Great Slave lake, lake La Martre and the Great Bear lake are situated in 
such a manner that a larger or smaller part of each lies within the 
Archaean shield, the remaining part within the Palaeozoic border. The 
boundary then reaches Coronation gulf, probably follows Simpson straits, 
and runs through Melville peninsula, but has not been determined farther 
to the south. The Palaeozoic border is known in East Melville, on the 
Nettilling, and at the upper end of Frobisher bay. It is not known to be 
eontinaed into Hudson strait. We find it again at Anticosti and in 
patehes on the very severely denuded region of the Saguenay. 

In the Arctic archipelago the border is formed of zones of Palaeozoic 
rocks* following one another in regular order and overlaid towards the 
north by patches of Mesozoic deposits. Towards the west, where it consists 
chiefly or exclusively of Devonian, the border sinks beneath the Cretaceous 
beds of the prairies. Towards the south it passes into the folding of the 
Cincinnati uplift and the folds of the Appalachians.' To the south-east 
it 18 bounded by the St. Lawrence, and the folded mountains of New 
Brunswick and Newfoundland. To the east and north-east an independent 
gneus range forms a boundary along the coast and runs from the strait 
of Belle Isle, on the coast of Labrador, to Hudson strait and Besolution 
island. Another independent lofty gneiss range, perhaps a continuation of 
that just mentioned, runs along the coast of Cumberland, is interrupted at 



44 THE OUTLINES OP THE ATLANTIC OCEAN [PARrra 

Home bay, and extends thence to cape Walter Bathursl From there lofty 
gneiss or granite mountains stretch through the east part of North Devon 
and along the coast of Ellesmere Land to cape Sabine in Smith sound. 

The numerous inlets, bays, and sounds of the Arctic archipelago lie on 
the flat-bedded Palaeozoic border, and the north coast of the mainland 
probably coincides for a long distance with the north edge of the shield 

2. The Baltic shield. The fundamental features of the part of North 
America just discussed are repeated in a remarkable manner in a part of 
northern Eurasia. Lapland and Finland are formed, like the Canadian 
shield^ of Archaean rocks, which were already folded before the Cambrian 
period : like it, they are surrounded by a curvilinear border of flatly bedded 
Palaeozoic sediments, owing their exposure chiefly to later erosion by ice, 
and finally, like it, they are covered by a vast labyrinth of small sheets of 
water, which are united by a maze of streams. In both cases a number 
of great bays and inland seas lie on the boundary region between the 
Archaean and Pala^eozoic rocks; the limit of the Palaeozoic girdle runs 
from the coast of Sweden near Oeland across the Baltic north of Gothland 
and north of Dago, through the gulf of Finland, quite close to its southern 
shore, then through lake Ladoga, also close to its southern shore, then, with 
a somewhat more oblique direction to the north, through the south part of 
lake Onega ; it reaches the sea not far from the town of Onega, probably 
crosses the peninsula which runs out to the north, and the gulf of 
Archangel, and then extends beneath the sea up to cape Voronov. 

Thus the Baltic sea, like Hudson bay, represents a partial submergence 
of the abraded Archaean platform, though in this case the submergence is 
less extensive ; the gulf of Finland, lakes Ladoga and Onega, and the bays 
of the White sea occupy a position similar to that of lakes Superior, 
Winnipeg, and the other oft-cited lakes of the American festoon, together 
with Coronation gulf and its continuations. To all these we may apply 
the observation, made long ago by Richardson with regard to the American 
lakes, that in each case without exception one part of the shore is Archaean, 
another part Palaeozoic, and that the lakes lying to the south of the 
boundary, on the Palaeozoic region itself, such as lake Peipus, lake Ilmen 
and others, correspond to Winnipegosis and its companions. 

The Palaeozoic girdle is distinguished in its western part and even 
south of the gulf of Finland by the fact that the several members of the 
Cambrian and Silurian systems rest upon one another in regular succession, 
so that as the distance increases the deposits become continually more 
recent ; but further to the north-east the red Devonian sandstone, coming 
from the south and east, encroaches more and more on the parallel Silurian 
zones, covering them out of sight one after the other, so that on the south 
shore of lake Ladoga none but their inner, that is older, members remain 
visible between the Devonian sandstone and the Archaean rocks, till finally 



cH.n] THE BAETIC SfflELD 45 

this sandstone comes directly in contact with the Archaean region, and on 
the coast of the North sea neither Cambrian nor Silurian beds are to be 
seen, but ozdy Archaean rocks and the red sandstone. 

This transgression of the Devonian sandstone has long been known : 
Hnrchison ^ has given a clear account of it ; and since I shall have to refer 
to it later, the following details may be introduced here. 

On the long island of Odand the strike, according to Dames, follows 
approximately the direction of the island itself, which is elongated to the 
north-north-east, so that the Cambrian beds lie on its west side, and to 
the east of these the lower Silurian stages succeed each other in regular 
order ^. Gothland belongs wholly to the upper Silurian; the various 
sabdiviaions of this system f oUow one another in ascending order towanls 
the south-east, and, as shown on the map drawn up by F. Schmidt, their 
boundaries run obliquely across the island to the north-east \ 

The detailed descriptions of Grewingk and F. Schmidt show that the 
strike, directed in Oeland to the north-north-east and in Gothland to the 
north-east, is bent round beneath the sea so completely that in Dago, Oesel, 
and Esthonia it runs from west to east parallel to the south shore of the 
gulf of Finland. The Cambrian deposits follow this coast, to the south 
they are succeeded by beds of lower Silurian, as in Oeland ; Dago and 
Oesel «kre formed of upper Silurian, as ia Oothland, and the highest Silurian 
beds are exposed in the south part of Oesel. The red Devonian sandstone 
forms the shores of the gulf of Riga, and at the north-east comer of this 
golf, near the bay of Peman, it overlies the upper Silurian ; the highest 
member of the upper Silurian, however, which is visible in the south of 
Oesel, seems to have already disappeared, and the red sandstone now 
advances farther and further to the north, concealing the Silurian, so 
Uiat between Saint Petersburg and Gatschina only the Cambrian and 
a narrow band of lower Silurian beds are to be seen ; at the entrance of 
the Sjas into lake Ladoga a Cambrian zone alone is visible beneath the red 
sandstone, and soon nothing but Archaean, rocks and red sandstone form 
the surface of the ground \ 

' R. I. Murchison, E. de Vemenil, and A. von E^ayserling, The Geology of Runia in 
Eaiope and the UnJ Mountains, 4to, 1845, I, pp. 41-49. 

' W. Dames, Geologische Reisenotizen aus Schweden; Zeitsch. deutsch. geol. Gesellsch., 
1881. XXXIII, pp. 415-483. 

3 F. Schmidt, Beitrftge zor Geologie der Insel Gotland ; Arch. Naturk. LIy-, Esth- and 
Knrlands, Dorpat, 1859, 1. Ser., Bd. II, pp. 403-464, map. 

* C. Giewingk, Erl&ntenmg zur zweiten Ausgabe der geognostischen Earte liv-, Esth- 
and Knrlands, Svo, Dorpat, 1879 (with map in FoL); also in Arch. Naturk. Dorpat, 
1- Ser., BcL Yin ; F. Schmidt, Revision des ostbaltischen silurischen Trilobiten nebst 
geognostisclier TJebersicht des ostbaltischen Siluigebietes, M^m. Acad. Imp. Sci. 
Stti^-P^nb., 1881, 7* s^r., XXX, pp. 55 et seq. ; and by the same, On the Silurian (and 
Ou&bzian) Strata of the Baltic Provinces of Russia, Quart Joum. Geol. Soc., 1882, 
iXXVin, pp. 514-535, map, pi. zxiii. 



46 THE OUTLINES OF THE ATLANTIC OCEAN [pakp nr 

In this region there are certainly some slight undulations in the 
Silurian, but the most marked of these appear to be nothing more than 
those accessory disturbances of dragging out or shattering, which 
accompany faults: generally speaking, the bedding of the Palaeozoic 
sediments is everywhere very flat. In striking contrast to this, the 
Archaean rocks are folded, and strike, as Inostranzeff has shown, along 
the great inland seas to the south-east, i.e. across « the course of the 
Palaeozoic girdle; their strike determines the trend of the peninsulas 
in the north part of lake Onega ^. 

On the south and south-e€U9t shore of this lake the predominant rock 
is the red sandstone, which as we have already seen occupies the eastern 
part of the gulf of Onega. On the little island of Ea-Ostrof , a few miles 
north of the town of Onega, Murchison observed granitic gneiss with 
vertical foliation*. 

Let us now turn our attention to the south of the Scandinavian 
peninsula. 

A glance at the map, might naturally lead us to suppose that Scania 
stands in an independent position. If the slightly curved coast of BohuB 
and Halland were directly joined to that of Blekinge, the south coast of 
Scandinavia would present a much more regular outline, and Scania seems 
to be added on as an appendage. Closer examination confirms this view. 
Scania is formed of fragments of a great plateau, broken up by fractures. 
It presents us with Mesozoic sediments which are elsewhere unknown in 
Scandinavia, although they once probably possessed a much wider 
extension towards the north. Apart from pre-Silurian and Silurian 
rocks we meet with red clay, usually assigned to the Eeuper, then coal- 
bearing sediments belonging to the Rhaetic and lias, the latter with 
marine beds, and finally the middle and upper Cretaceous. In one 
locality the coal-bearing series abuts against the gneiss, in another against 
the Silurian ; the Cretaceous rests sometimes on the Lias, sometimes abuts 
aficainst the Silurian, at others overlies the giieiss, and as E. Erdmann has 
sbowB the whole region is traversed by greTt longitudinal fractures which 
run from north-west to south-east; along these the whole country has 
been let down irregularly, with the formation of troughs and horsta 
The investigations of Nathorst have shown fui4)her that this subsidence 
has taken place at difierent times, and that it is possible to classify the 
fractures according to their age '. 

^ A. A. Inostranzeff, Carte g^ologiqne de la partie septentrionale da Goayemement 
Olonetz ; Mat^r. Gr^oL Rnss., 1877, VII, pi. ii at passim. 

* Murchison, The Geology of Russia in Europe and the Ural Mountains, pp. 21, 22. 

' E. Erdmann, Description de la formation carbonif^re de la Scanie, 4to, 1878 (in the 
publications of the Royal Swedish Greological Institution) ; Herr Erdmann has kindly 
furnished me with more recent observations on the fault lines ; G. Nathorst, Till Mgan 



can] TABULAR FRAOTURES OF SCANIA 47 

We now recognize Hallands ka with HallandH Vaderci, north of the 
Skdder Vik, and the Knllen, south of this bay, as two horsts, rmuiing out 
to tiM nortii-west ; Bomele Khnt in the interior of the country is a sbnilar 
bont, and the projection of the land to the eoath-eaat between Ystad and 
CSmtgiahamn is determined in like manner by the bactnres, which ran to 
tite south-east We can distinguish among these fractures those which 
mx^ produced after the Silurian and before the Keuper from others which 



Fis. 6. SaaUa, afl«r G. Hathant. 
A Archaean; * Carobrisa and Silurian; k Keuper; rh Bhaetia and Liaa ; cr Cretaoeona; 
> Diabaaa ; B— alt ahown in black. Strong line*, definitely aaoertained bnlta ; broken Unea, 
co^jectoial bnlta. 

were formed after the Keuper and before the Cretaceous, and these from 
yet otiierB which were formed after the Cretaceous ; and thus the depodtion 
of the Cretaceous on the gneiss, i. e. on an . old exposed horst, finds an 
explanation. Nathorst has shown that the horst of the Kullen, for 
example, as well as its probable continuation the SOder Asen, is bounded 

«M de Sk&nikft dulocationernas Uder, Geol. Foren. Stockh. TOib., 1887, IX, pp. 74-180, 
■^iL One of the tnunrene fJBctniei ii clearly fi^i«d b; Haacheconie in Zeitochr. f. 
Beig-, Htltt a. Sftlinenw., 1875, XXUI, pUt« b in teit 



n 



48 THE OUTLINES OF THE ATLANTIC OCEAN [PARrin 

on the south by a fracture which is older than the Trias, and on the north 
by another which is younger than the Cretaceous, and in this country, 
which has not been subjected to any folding since the Cambrian period, 
we have a new and an instructive example of the breaking down of 
a table-land accomplished piece by piece and at different times. A few- 
short transverse fractures, running to the north-north-east, also occur. 
The older diabase dykes follow the master fractures to the south-east, 
and the younger basaltic dykes appear to correspond to the north-north- 
east strike of the transverse fissures. 

Bomholm is the continuation of this fractured table-land or rather of 
the horst, formed of gneiss and Palaeozoic sediments, which projects at 
Melby Asen and near Cimbrishamn. In the south-west part of Bomholm 
the subsided coal-bearing series of Scania is visible, together vdth down- 
thrown patches o£ Cretaceous ^. 

Now that we have separated off Scania and Bomholm let us cast 
a glance over the great peninsula of Scandinavia. 

A lofty range of mountains rises in the west' facing the Atlantic 
Ocean; to the east and south next the Baltic and the Skager Back 
extends a broad tract, lying at a low level in the whole of the south part of 
the peninsula, and consisting chiefly of Archaean rocks. The boundary 
between the mountains in the west and the Archaean land in the east is 
marked by a great scarp, which may be followed from Stavanger in 
lat. bd'' N. into the northern part of the district of Tromso in lat. 70'' N. : 
it describes an irregular and frequently interrupted line, but on the whole 
follows the main direction of the peninsula. Although the mountainous 
region consists in great part of folded ranges which mostly run out to the 
Atlantic coast, cutting it obliquely, yet much of the eastern slope presents 
us with beds which lie quite flat, and facing it on the east there rise from 
the Archaean region isolated mountain masses, formed of the same hori- 
zontal beds, which must evidently be regarded as outliers of the great 
escarpment. Thus the face of the escarpment certainly does not correspond 
as a whole with a continuous fracture ; over the greater part of its course 
it is without doubt a denudation boundary, and as Pettersen rightly 
pointed out, when describing its northern part many years ago, the 
mountains in that region must once have extended beyond the escarpment 
much further to the east than they do at present. 

The Archaean region is bounded on the east by that flat-bedded 
Silurian belt which borders the peninsula west of Oeland, and thence 
runs as we have already seen beneath the Baltic to the gulf of Finland. 

■ 

^ K, v. Seebach, Beitr&ge znr Geologie von Bornholm, Zeitsch. deutsch. geol. Ges., 1865, 
XVII, pp. 838-347, pi. viii ; M. Jespenen, Bidrag til Bornholm's Geotektonik, Nat. Tidsp 
^Obenh., 1867, 3. R., V, pp. 33-52, pi. yii ; Nathorst, Till fifigan om de Sk&nska disloca- 
tionemas &lder, Geol. FOren. Stockh. FOrh., 1887, IX, pp. 116 et seq. 



CH. n] THE SILURIAN OF CHRISTIANA 49 

It indudeB the south of Norway and most of Sweden. It rises towards 
the north, where it is covered by dense forest ; in the south it bears the 
great lakes; everywhere it shows traces of the grinding action of the 
ice. It was formerly covered by a continuous sheet of Silurian beds, 
and the isolated fragments of this sheet which have escaped destruction 
doubtless owe their preservation to subsidence. The most important of 

■ 

the downthrown patches is the remarkable band of Silurian deposits, more 
than 200 kilometers long, which runs from the east slope of the mountains 
to the south-south-west across lake Mjosen and Christiania, and then down 
the west side of the Qord ; its middle part is often spoken of as the Silurian 
region of Christiania. 

This band has indeed experienced strange vicissitudea Considerable 
parts of it first underwent folding and subsequently abrasion ; then it was 
broken up into segments and subsided; later or perhaps simultaneously, 
copious injections of a red granite, the Drammen granite, welled up into it ; 
then it was traversed by more recent eruptive dykes, and finally cut up by 
a number of smaller trough faults, which form the existing transverse 
valleys \ 

In the north, on lake Mjosen, the folding is very intense, and over- 
folding also occurs. At Christiania the folding is equally intense ; south 
and south-west of Christiania the coasts of the fjord afford many well- 
known examples of structural features, such as intrusions of more recent 
granite in the form of laccoUtes, with an altered roof of Silurian lim^tone ; 
fractures bringing the Silurian beds against the ancient gneiss ; and finally 
horizontal displacements of orographic blocks. On this coast also, to the 
south-west, are found the admirable examples of small trough faults 
which have been described by Ejerulf. Brogger has given an extremely 
instructive account, describing in detail the movements of the mass and 
the history of its injection K 

Still further to the south, between Skien and Langesund, folding has 

' T. ^emlf, Dislokationerne i Krittiania dalen ; Nyt Mag. Naturvid. Christiania, 1883, 
iiVlII, pp. 79-88 and 171-197, and in many other publications. 

* W. C. BrOgger, Die siloriBchen Etagen 2 und 3 im Kristianiagebiet und auf Egger, 
Onxrerntftts-Programm f&r das 2. Semester 1882, 8vo, Eristiania, 1882 ; and in particular 
by the same, Ueber die Bildungsgeschichte des Eristiania Qords, Nyt Mag. Naturvid., 
1886, XXX. pp. 96-244, map ; farther, Om Eristiania Qordens Dannelae, Naturen, 1886, 
BO. 7, 8 ; also A. Penck, Ueber einige Eontaktgesteine des Kristiania-Silurbeckens, Nyt 
Mag. Natorrid., 1881, XXV, pp. 62-82 ; E. Reyer, Vier Ausflflge in die Emptivmassen bei 
Chiistiaiiia, Jafarb. k.-k. geol. Reichsanst., 1880, XXX, pp. 27-42, &c. I have myself had 
an opportunity of visiting with Herr L. Burgerstein some of the most characteristic 
points of contact of the Drammen granite with the Silurian deposits, of collecting at the 
s«iie of contact fragments of Halysites converted into Yesuvian,' and of convincing 
myself as to the correctness of Ejerulf s description of the Drammen granite. Br0gger*s 
investigatiobs are among the most detailed which we possess on the formation of a great 
trough. 

n E 



50 THE OUTLINES OP THE ATLANTIC OCEAN [part in 

disappeared. The narrow strip of Silurian rocks rests towards the west on 
gneiss and towards the east it sinks with a rapidly increasing dip, that is 
as though by flexure, beneath a very extensive mass of syenite. This 
narrow strip is also traversed, as shown by Brogger, by longitudinal and 
transverse faults and by many intrusive, dykes ; the variotis segments have 
been let down to different depths. Brogger has at the same time shown 
that the subsidence of these segments and the ascent of the eruptive rocks 
are probably connected phenomena \ Since, however, the island of Nord 
Eoster also belongs to the Silurian, it is possible that the whole coast up 
to the faults of Scania will prove to be a continuation of the eastern 
fracture. 

To the east of the great downthrown zone, fragments of ancient 
sedimentary masses are also faulted in ; these occur on the south-east side 
of lake Wenem, between this lake and lake Wettem, on the east shore of 
lake Wettem, north of this lake, and in other places ; Nathorst is even of 
opinion that the line of demarcation between the gneiss and the granite, 
so sharply defined in south Sweden, represents a great pre-Silurian fault 
The supposed fault would run from the neighbourhood of Solvesborg, on 
the south coast, towards the north, crossing obliquely the southernmost 
part of lake Wettem, and touching the east side of lake Wenem it would 
then proceed, with a slight deviation towards the north-north-west, up to 
the high mountain ranges. It would extend from lat. 56"" to lat. 61 "* N., 
and the downthrown part would be the gneiss forming the west limb. 
The intrusions of hyperite, which make their appearance in the gneiss close 
to the supposed fault-line, might then be a direct consequence of the 
subsidence itself '. 

The view that great subsidences have affected this Archaean region in 
andent times appears to be rapidly gaining ground among Scandinavian 
geologists, and Svedmark has drawn up a whole system of conjectural 
lines of fracture in the region extending north of Stockholm as far as the 
sea of Aland, which is distinguished by its extraordinary depth \ — 

The mountain region, which is to no small extent a true table-land, 
presents a fundamentally different character. 

From Eragero in the east to Stavanger in the west, the whole south 
coast of Norway is formed of gneiss and granite, and the same rocka 
passing through Christiansand form all the country far into the interior. 
This is the western part of the Archaean region. Above it, extending 

» 

* W. C. BrOgger, Spaltenverwerfangen in der Gegend Lsuigesond-Skien ; Nyt Ma^. 
Natunid., 1884, XXYIII, pp. 258>419, map. 

' G. Nathont, Ett fSnOk att fOrklara onaken till den skarpa grftnsen mellan eOdia 
Sveriges vestra och Qrtia urterritorinm ; GeoL FOren. Stockh. FOrh., 1886, YIII, pp. 95-102. 

' E. Svedmark, Orografiska stadier inom Roslagen ; Geol. FOren. Stookh. FOrh., 1887, 
IX, pp. 188-210, map. 



cH.n] LANGPJELD, THE SPARAGMITE 51 

bom Stavanger to the north-east, a steep scarp rises which runs in and 
oat with an irregular lobate outline; this is the edge of the extensive 
Langfjeld and at the same time the beginning of the great eastern slope. 
The most striking feature presented by this scarp is the contrast it offers 
to the r^ons lying more to the east in the thickness attained by the 
Cambrian system, owing to the intercalation of great masses of quajrtzite 
and schists. Tellef Dahll has examined a part of it to the north-east of 
Stavanger; he found that the escarpment was formed of approximately 
horiaontal beds of quartzite and lustrous schists, and in the Huulberg, 
where the boundaries of the three southern dioceses of Norway unite, 
he discovered in the lower part of this series a bed with Dictyograpbus 
0abeiUfor7ni8 (ssDictytynevna wcidle) ; the whole series is therefore assigned 
to the * primordial * stage. ' On the Hallingskarve, near the upper end of 
the Hardanger fjord, a recent granite is found, according to the same 
observer, overlying this series. In other places superimposed masses of 
mica schist and hornblende schist are described as piled high above it \ 

The flat-bedded series of schists and quartzites is continued to the 
north-east over a part of the Vidden, through Hallingdal and into the 
Gudfarandsdal ; it forms also some isolated masses which face the east 
border of the mountains. The quartzite of the great ranges must be 
regarded as a higher member of the series, but in this r^on also mica 
schists, augen-gneiss and hornblende schists are said to occur high on the 
summits. The arrangement of the beds, however, is not sufficiently 
known ^. 

To the north of this region rise the rugged mountains of the Jotun 
Qeld, formed of gabbro. The sketch-map published by Beusch gives the 
impression that this series of great masses of gabbro marks the boundary 
between the flat-bedded eastern part of the range and the western folded 
part*. 

The flat-bedded girdle soon changes its character, and we reach a region 
which it is particularly difficult to describe in detaiL I follow the data 
given in a series of masterly descriptions by Ejerulf ^. 

On the side of the Gudbrandsdal the flat-bedded series presents itself 
as an alternation of schists and blue quartzites ; hitherto they have rested 
on the Archaean rocks ; now a mighty series of feldspar-bearing sandstone, 
the ' sparagmite * of Norwegian geologists, is inserted beneath them ; the 

' Tellaf Bahll, Ueber die Greologie Tellemarkens, translated into German by 
W. Ghiirtopherten, 4to, Chrisiiania, 1860, pp. 16-19. 

' IQemlf, Die Oeologie dea eadlidien und mittleren Norwegen, p. 184. 

' Hans H. Renach, Die fosnlienftthrenden kxyftallinischen Schiefer von Bergen in 
Norwegen, Gennan translation by R. Baldaof, 8to, Leipzig, 1883, p. 5. 

« Of these I will only mention Geologisk Oversigtskart over det sydlige Norge, foL 
1878, sad T. ^emli; Die Geologie des sadlichen und mittleren Norwegen, German 
tnaslatioB bj Gnrlt, 8vo, Bonn, 1880. 

E 2 



52 THE OUTLINES OF THE ATLANTIC OCEAN [pabt m 

upper part of this series yields Paradoxides and may therefore be definitely 
assigned to the Cambrian system. . In the lower part of the sparagmite 
lies the unfossiliferous zone of the Birid limestone. But while the 
sparagmite with the Birid limestone lies beneath the schists and blue 
quartzite, the richly fossiliferous series of the lower Silurian lies above 
the stage of the blue quartzite and, as it appears, unconformably, so that 
the Orthoceras limestone, for example, rests on different stages of the 
Cambrian system. At lake Mjosen the lower Silurian is present in 
considerable thickness, and it is from the neighbourhood of this lake that 
the long strip of faulted-down Silurian branches off to the south-south- 
west, and, as we have already seen, reaches the sea by way of Christiania, 
thus dividing the Archaean region into two parts. The thick accumula* 
tions of the Cambrian, with which we have just become acquainted aa the 
blue quartzite and the sparagmite stage, do not accompany this band. 

On the Herjehogna, a peak 8,800 feet in height, situated in Dalame . 
at about lat. 61'' SO' N., the rampart of quartzite crosses the Swedish 
frontier. It marks the eastern edge of the mountains, but not, in this region, 
the east boundary of the sedimentary beds which are superposed on the 
Archaean foundation, for in front of it and emerging from beneath its 
foot the continuation of the Norwegian sparagmite, here called the Dala- 
sandstone, makes its appearance and extends at no great height over 
a great part of southern Dalame. 

For a knowledge of these regions and of those extending to the north 
through Herje&dalen and Jemtland to lat. 64"* N., the works by Tomebohm 
published in 1878 are the chief source of information, and it is the results 
obtained by this indefatigable observer which will now be my guide \ 

The quartzite which forms the great rampart, here called the Wemdal 
quartzite, is the continuation of the high-mountain quartzite of Norway ; 
the rampart bounds the mountainous region. At its foot, a little north 
of lat. 62'' N., the sparagmite disappears, yet the rampart does not rest 
directly on the Archaean rocks ; for close to its foot a series of Silurian 
beds is intercalated, in particular the Orthoceras limestone ; this accom- 
panies the quartzite rampart into the neighbourhood of Arsames Eapell, 
in about lat. 62* 45' N., forming a narrow belt, and according to Tome- 

^ A. E. TOrnebohm, Ueber die Qeognode der schwediechen Hochgebirge; Bihang 
svenska Vet. Akad. Handl., I, no. 12, 1878, map ; Sver. geol. UndersCkn., ser. C, no. 9. 
Part of this previously appeared in En geognostisk Profil OfVer den Skandinaviaka 
i^&Uryggen mellan Oatersund och Levanger ; by the same, OfVerg. E. Vet. Akad. FOrh., 
1872, and Sver. geol. UndersOkn., ser. G, no. 6, 1872. TOmebobm has distinguished 
in the high mountain beds two groups, the Seve and the Eoli; but subsequent 
observers do not seem to have considered these new designations necessary. A good 
general geological map of the region around the Stor Sjdn is contained in A. G.Hdgbom, 
Glaciala och Petrografiska Jakttogelser i Jemtlands L&n; Sver. geol. UndersOkn*, ser. C, 
no. 70, 4to, 1885. 



can] JEMTLAND 53 

bohm underlying the qoartzite. From this point the rampart recedes 
yery far towards the north-west, only advancing again in the north, 
nod it thus forms an irregular arc within which the Silurian beds are 
distributed with horizontal bedding. On these tabular masses of Silurian 
lies the great lake of Stor Sjon. Towards the east these masses rest on 
the Archaean rocks, which here, however, attain a greater elevation than 
in the south, and even rise higher than the Silurian table-lands. 

We now scale the quartzite rampart and reach the mountainous region, 
OQ which several summits attain a height of 5,000 to 6,000 feet ; a con- 
siderable part lies above the limit of forest growth, i«e. over 2,800 feet. 
Here a lower group may be distinguished, consisting of quartzite and 
mi^ty maases of crystalline schists, namely mica schists, hornblende 
sehistSyaad even gneiss; and an upper group which comprises semi-crystal- 
line day slates and likewise hornblende schists. Strange though it may 
seem to encounter gneiss and hornblende schists high above the Silurian 
in normal superposition, yet we must remember that a precisely similar 
phenomenon has been observed by Norwegian investigators in the south 
part of the great mountain border. In short, the succession of rocks as 
^ven by Tomebohm is as follows : older granite (Archaean) ; on this, 
Dala-sandstone (sparagmite, only in the south); Silurian in the north- 
east up to the horizon of the Pentamerus limestone ; then Wemdal 
quartzite, which forms the great rampart above the Silurian, and above 
this another mighty series of highly crystalline schists. 

The conclusions arrived at by Tomebohm have been disputed. 
Sveuonius, as the result of a series of detailed surveys, considers that 
the narrow belt of Silurian deposits at the foot of the quartzite rampart, 
and with it the expanse of Silurian about Stor Sjon, do not really underlie 
the quartzite ; the quartzite is certainly traversed by several folds, while 
the Silurian shows absolutely no signs of folding; thus the quartzite 
rampart represents a Silurian shore, and the Silurian was deposited 
against it, as well as in the pre-Silurian valleys ^ 

This view, however, which accords so well vdth the older conceptions 
of the great antiquity of the crystalline schists, was destined soon to 
nndeigo great modification, chiefly owing to the continued labours of the 
same conscientious investigator. The most important of the fresh obser- 
vations was the discovery of Silxudan fossils at many localities within the 
great ranges themselves. The imtiring zeal of Scandinavian geologists 

' F. SvenoniiiB, Till Fr&gan om FOrh&llandet mellan * Wem8darft<)nartBiten ' och 
sUaiiika Formationen inom 80dra Delen af Jftrntlands Lftn ; Sver. geol. UndenOkn., ser. C, 
ao. 49, 8to ; also Qhen. E. Yet. Akad. FOrh., 1881, no. 10. HOgbom has since shown 
tliat parts of the east border of the Silurian region are fiiulted down into the Archaean 
fmiiidation. By th« same, Om fOrkastnings-breccior vid den Jftmtlandska Silurforma- 
tbnens Ostra gr&ns ; Geol. FOren. Siockh. FOrh., 1886, VIII, p. 352. 



64 THE OUTLINES OF THE ATLANTIC OCEAN [PABrm 

will certainly succeed in dissipating at last the doubts which at present 
exist as to the relation between the Silurian table-lands of the Stor Sjdn 
and the high mountain ranges. Let us now turn our attention to the 
north while continuing to follow the eastern boundary of the mountains. 

A little way to the north, very nearly in lat. 64° W N., where the 
Sjougdelv, coming from the west, flows into the Tasjon, Svenonius met 
with fossiliferous -Cambrian deposits. These are folded with the quartzite 
and lie in a synclinal of this rock ^. In the northern part of Jemtland and 
in Westerbotten, a new element now makes its appearance in the mountain 
border, namely, a very long series of large and small masses of olivine 
rock and serpentine, which are intercalated like sills with the schists of 
the high mountains. 

East of Grong, in the midst of the Norwegian mountains, in lat. 64'' N., 
Hauan met with a great mass of anorthite olivine rock \ the beginning, 
apparently, of an extremely long band, which follows the general strike 
of the mountains from here towards the north-north-east, crosses the 
Swedish frontier, and becomes visible on the east slope. Many isolated 
exposures had been known from comparatively early times; Svenonius 
has shown their continuity. The largest masses are the Bodfjallet on the 
Stor Bl&sjon near the headwaters of the Sjougdelv, mentioned above, and 
the mountains of Graipies and Oma, which lie in close proximity to one 
another north of Fatmomak (lat 65"" N.). The series is continued much 
further still : moimtains of olivine occur even beyond lat. 67* N. to the 
south-west, west, and north-west of Evikkjokk. 

It was in this region also that Svenonius was quite recently led to 
assign a ^ post- Azoic' age to these mighty, high-mountain formations, 
owing to the discovery of fossils, particularly of a bed containing Hyolithes 
in the Paije Sartajaur and of Encrinite joints in the high mountains. The 
same observer has had the kindness to inform me that he has now found 
the Hyolithes beds in five or six places in Norbottens Lappmark, but not 
yet in Jemtland, and, on the other hand, he has met with fragments 
of Crinoids in the extreme northern part of Jemtland and at scnne 
localities in Westerbottens Lappmark, in the calc-mica schist of the high 
mountains. The beds of Hyolithes appear to underlie the mountains and 
form a belt along their eastern boundary. 

^ Svenonius, Om 'Seve gruppen' i nordligaste J&mtland och Angennanland, eamt 
dess FOrh&llende till fosdlfOrande Lager; Sver. geol. UndenOkn., ser. C, no. 45, Svo, and 
Geol. FOren. Stockh. FOrh., 1881, Y, pp. 484-497, pL zz, xzL 

* Ejerulf, Geologie des sfidlichen Norwegen, p. 272. 

' Svenonius, Om Olivinstens och Serpentin fdrekomster i Norrland, Sver. geol. Under- 
sQkn., ser. C, no. 56, and Geol. FOren. Stockh. FOrh., 1883, YI ; by the same, Stndier vid 
Svenska Joklar, Sver. geoL Undersakn., ser. C, no. 61, and Geol. F6ren. Stockh. F6rh., 
1884, YII^ pp. 5-38, and Nya olivinstens f^rekomster i Norrland, tom. cit, ser.C, no. 61, 
and Geol. FCren. Stockh. FOrh., 1885, YII, pp. 201-2ia 



can] LAPPMARK 55 

We now enter a r^on which has been made known in its general out* 
lines by Hummers sorvey, and by a section drawn by K Pettersen across 
the peninsula^ from north-west to south-east, i.e. from Saltdalen to Pite&^. 

The peninsula consists here, according to Pettersen, of the following 
partSL The gaH of Bothnia is bordered on the coast by a narrow belt of 
lowland, fonned of gneiss in steeply upturned beds. The country then 
suddeoly rises towards the interior to a height of about 1»100 feet, and 
from here there extends a monotonous granitic highland, for the greater 
part covered by forest and about 200 kilometers broad. It rises very 
slowly on the west towards the zone of the great lakes, so that the water- 
level of the Saedva for example, west of the Homafvan, lies 1,420 feet 
above the sea. Towards the lakes some degree of variety replaces the 
monotony of the east; isolated ranges rise some hundreds of feet above 
the surrounding country. Finally, at the west end of the Homafvan and 
between this great lake and the Laiselv, which follows on the south, we 
reach the great rampart^ which still marks the east edge of the mountain 
highland. Its highest ridge lies at a height of about 2fiOO feet* It 
appears from Hummel's observations that the northern continuation of 
tiie rampart is crossed by several of the great lakes; this is the case to 
the east of Kvikkjokk, and the Tome& Trask lies in such a manner that 
afanoet its whole western part is let down into the mountain highland, 
while the larger^ eastern part lies outside it. 

Above the rampart^ a little way to the south, there ris^ the isolated 
peak of Peljakaisse (1,064 meters), which lies 275 kilometers from the 
golf of Bothnia, and 187 kilometers from the Atlantic Ocean; and now 
the snowy peaks crowd higher and higher towards Sulitjelma (1,875 
meters) and the far higher summits in the north, namely Sarjektj&kko 
(2,128 meters) in Lule&, and Eebnekaisse (2,156 meters) in Tome& 
Lappmark *. 

The rampart itself from north Jemtland up through Lappmark con- 
sists, so far as yet known, of a mighty series of strata lying horizontal. 
Pettersen distinguishes a group of beds at the foot of the rampart, which 

. ^ David Hummel in Underd&nig BerfttteLse om en p& n&dig Befallning &r 1875 
f<Dretagen UndenOkning af Malmfyndigheter inom Gellivare och Jukkaig&ryi Socknar af 
Nonbottens lAn, Sver. geol. UndersOkn., 4to, 1877; see also pi. i in this memoir, 
Qeologiflk Ofvenigtskaita Ofver den k&nda Delen af NorrbottenB L&n, and pi. ii, 
Geologisk Karta Ofver en del af Tome och Lale Lappmarker; Karl Pettersen, Det 
aoidlige Sveriges och Noiges Greologi ; med et geologisk Profil over den Skandinaviske 
Haho fra Saltdalen til Pite&, in Lie, Mflller and Sars, Archiv. Math. Naturvid. Christiania, 
1878, III. 

' Srenonins, N&gra ord om Svenska Lappland, in the publication called ' Heidnin,' 
Febmaiy and March, 1885, pp. 26-83. The altitudea, as Herr Svenonius informs me, 
we taken ftom 'Koirbottens l&ns Ekonomiska Eartwerk' and rest on trustworthy 
aessorements. 



56 THE OUTLINES OF THE ATLANTIC OCEAN [PARTiir 

he calls the Dividal group, after a valley in Tromso 8tift» south of the 
Bals fjord ; it consists of fine-grained schists, red» green or grey in colour, 
of quartzite and quartz schists, and corresponds to that zone in which 
Svenonius found Hyolithes ; we assign it to the Cambrian system. Above 
this group lies the group of the Tromso mica schists, composed of mica 
schists and quartzite. 

In order to discuss the further continuation of the rampart beyond 
the Tome& Trask, so far as it is known, we must first consider the main 
features in the structure of northern Norway. Pettersen's extensive 
investigations and the geological map by Tellef Dahll are the most im- 
portant sources of information ^. 

First of all, it must be borne in mind that the great chain of islands 
which borders the mainland on the west, the Lofoten and the Westeraalen, 
is formed of gneiss and granite, evidently of very great age ; and thus 
these islands probably exhibit the oldest rocks which exist in the mountain 
ranges of the north ^. We shall soon see that the Hebrides similarly 
present us with the most ancient rocks of Scotland. In the north-east 
part of the northernmost of these islands, at Bamsaa, in the island of 
Ando, there lies a small patch of more recent sediments. At one time 
coal mines were worked here, now said to be filled up. Some slabs of 
sandstone filled with valves of Aucella, to be seen in the museum at 
TromsOi show that these sediments are those Mesozoic deposits of the 
north which once extended from Spitzbergen to Ando^. This isolated 
patch lies close to the sea-coast, and although the data as tathe lie of the 
beds are very incomplete, yet the similar patches in Scotland show that 
we probably have here a faulted-down fragment, preserved as the last 
remains of a deposit which formerly extended over a wide area. 

Another element of particular importance in the structure of the 
country is the lofty and extremely rugged mountain range, which strikes, 
to the north-north-east and separates the Lyngen fjord from the Ulfs fjord 

^ Tellef Dahll, Geologisk Kart over det Nordlige Norge, mit Bistand af Gomeliuaaen, 
Hjortdahl, Lassen og C. Pettersen, foL, Christiania, 1886-1879: Pettenen's later 
treatises are contained partly in the TromsO Museum's Aarshefber, partly in Arch. Math. 
Naturvid. of Lie, Mflller, and Sars. Pettersen distinguishes three groups of stratified 
formations : Dividal group, TromsG mica schist, and Bals Qord group ; the latter has 
recently been regarded as older than the Tromsd mica schist ; since however this question 
is not of primary importance for the subjects treated here, I have not quoted, besides 
the Dividal group, any other members in the table-land, or, as it is usually called in 
Norway, the high mountain series; cf. Pettersen, BalsQordgruppens plads i den 
geologiske Folger&kke, TromsO l^^useum's Aarshefter, VL 1883, pp. 87-97. Several of 
the most important rocks of these regions are described by Philippson in Verb, naturh. 
Ver. preuss. Rheinl., 1883, XL, Sitzungsber., pp. 190-210. 

' Pettersen, Lofoten og Yesteraalen : Arch. Math. Naturvid., 1881, V, map. 

' Among the plant remains of Andd are two species of Pinus which 0. Heer har^ 
compared with those of Spitzbergen. 



CH. n] LYNGEN 57 

oa the west. Its black, terribly steep peaks rise to a height of about 
IJBOO meters above the sea ; glaciers descend between them ; the contrast 
wiih the gentler mountains of the surrounding country is extremely 
striking. The outlines recall, by their savage aspect, the tonalite crags 
of the Presanella in Tyrol. This range, which may be clearly seen on the 
map by Tellef Dahll, consists of gabbro and serpentine, and resembles to 
some extent the great 'cicatrice' of gabbro which forms the Jotun 
mountains in the south. 

Not far from the place where the upper extremities of the Bals fjord 
and the Lygen fjord approach each other in the south, rises the steep pyra- 
mid of the Pig-tind, which is a part of the Lyngen range ; a long, abrupt, 
but low ridge of gabbro, which strikes south of the Bals fjord obliquely 
to the west-south-west through the great valley of the Maalselv, may 
perhaps be considered as a continuation of this range, bent into the strike 
of the mountains ; but it is hard to say. Even much further to the west- 
south-west, e.g. at the foot of the Istind in Bardo, numeibus injections 
of edogite and gabbro are visible, as though a zone of fracture and 
intrusion ran from Amo in the north through Lyngen, and then parallel 
to the main strike across the vaUey of Bardo. 

The rocks which appear on the west of this zone towards the islands 
and in the islands themselves, up to the ancient gneiss, exhibit many dis- 
turbances, but in the interior of the Bals fjord, as well as towards Malanga, 
the wellnstratified mountain masses assume more and more the form of 
broad blocks, sometimes gently inclined, sometunes lying nearly hori- 
aontal. Violent local folding is present, but the mountain structure, as 
a whole, is not that of a folded range. 

The country lying to the east of the zone of intrusions, or the zone of 
the Lyngen gabbro, forms a continuous zone of stratified rocks extending 
to the east border of the mountains, i.e. to the great rampart. This is the 
zone of the high mountains, properly speaking. It is composed of clay 
dates, quartzite, and a later mica schist. 

The desire to become acquainted with one of the valley systems which 
open into the northern fjords, as far as the watershed, led me to visit these 
mountains, and the abundant directions and explanations with which 
Herr Pettersen kindly furnished me, enabled me in a short time to obtain 
some insight into their peculiar structure. Dr. Leo Burgerstein did me 
the favour of acoompanjdng me ^ 

^ I owe particular thanks to my companion, Dr. Leo Burgerstein, who, with true self- 
nciifice, strove to reduce for me, as far as possible, the inconveniences of such a journey, 
Mud thus enabled me to devote my whole time to my work. Most of the localities 
mentioned here may be seen on Pettersen*s general map of the south part of TromsO 
dkbiet, in his Geologiske UndersOgelser in den Troms^J Amt og tilgrsdndsende Dele af 
Notdlands Amt ; K. Norsk. Vid. Selsk. Skr., 1874, VII, pp. 261-444 and pL 



58 THE OUTLINES OF THE ATLANTIC OCEAN [pabtih 

Travelling southwards from the Malangen fjord to the lake of TomeU 
we reached the plain of the Stromsmoen (lat. 68* 48' N. ; 61*5 meters) ; 
and here, jnst where the Sordal, coming from the south, opens into the 
valley of Bardo, there rose before us a marvellous scene. Lofty monn* 
tains of horizontally stratified quartzite surround the plain, and reach 
a height of as much as 2,000 feet above the valley bottom : Borgsklattea 
in the east, BjdmeQeld and StorQeld in the west, and the wedge-diaped 
Bubben, which separates SSrdal from the upper Bardo, in the south* 
Bound about the plain there runs at a unifonn level from one mountain 
mass to another a narrow band of startling whiteness, looking as though 
it had just received a coating of whitewash. This is a bed of limestone 
which, intercalated with the dark quartzites, not only renders obvious 



In the jiisUnee the gabbro 
of Lyngen 



HOlltind 




N9rth. 



Fxo. 6. View /ram the Omancarre, south (if the BaUJjord^ 



the horizontal bedding, but also shows that this valley has been produced 
solely by erosion '. 

From here we continued our journey 25 kilometers further to the 
south through the S5rdal to the frontier, always among mountains 
formed of flat-lying beds. In its upper course the stream has cut for 
itself an inaccessible ravine, and we now ascend towards the south-east 
the ancient glacier bed of the Stagenuni (782 meters). Here we find 
ourselves standing above the quartzite, on coarsely foliated mica schists 
of a dark-grey or pinchbeck-brown colour. The great sheet of water 
which forms the north-west part of the Tome& Trask now lies spread 
out before us in incomparable beauty. On our right the opposite shore 
is bordered by a tongue of flat woodland, corresponding with the mouth 
of the Nuorajoki; further away to the left, the steep walls of the moun- 

* This white band has also been obBerved by Pettenen, Notiser TodrOrende den 
nordnonke Qeldbjrgning ; Geol. F6ren. Stockh. FOrh., 1886, YIII, p. 466. 



CH. n] 



OMASVARRE 



59 



tam Abesko advance cloae to the water's edge, and its broad but noble 
forms are mirrored in the lake. In the background there rise one above 
the other the white rounded, or dome-like summits of the mountains of 
Tome& Lappland. All the country visible at a little distance — ^the cubical 
smnmits of the west side of the Sordal, such as Spikalomi, and the equally 
high mountains of the east side, such as the Etnamjoeki peak, which 
belongs to the great group of the Duoddarats — is composed throughout 
of horizontal beds up to a height of more than 3,000 feet above the sea- 
level. We are thus standing on a fragment of true table-land, and 
descending to the lake we find its extremity divided by a projecting reef 
of qoartzite with the bedding equally horizonal. The great rampart how- 
ever lies, as Hummel has shown, still further to the east. 



Hatteyarre 




IIJIU Frisk 
Trwu6 fltifl (lat. 69* 8' N.) ioirards the east. 



SoQtb 



The structure of the mountain land lying south of the Bals fjord may 
be seen from the sketch (Fig. 6), made from the broad ridge of the Omas- 
varre (670 meters, lat. 69^ 8^ N.), situated 11 kilometers south of the 
end of this fjord. It was drawn looking to the east, and on the left, 
i.e. towards the north, a few of the peaks of the Lyngen range are visible. 

These peaks excepted, all the mountains are well stratified; the 
colouring, the alternation of schist and quartzite, and the long lines of 
snow resting on the ledges, all combine to bring out the lie of the beds 
in clearest detail. On the right, not only the broad mass of the Buosta- 
Qeld (1,672 meters), just outside the limits of the sketch, but all the 
mountain land far to the south is formed of horizontal beds. The first 
mountain represented on the right, the southernmost member of the group 
of the Hattevarre, has the form of a regular pyramid, yet it nevertheless 
equally consists of horizontal beds. The beds of the principal mass of 
the Hattevarre are also nearly horizontal, but towards the Bismaalstind 



n 



60 THE OUTLINES OP THE ATLANTIC OCEAN [part hi 

inclination sets in and increases towards the north ; in the Bismaalstind 
itself the dip is already very marked, and in the HoUtind it has 
become so steep that the north flank is formed by a single bedding plane ; 
this plane plunges into the Bals fjord, and the mountains which we see 
further away towards the Lyngen range also form part of the side of the 
fjord. There is nowhere any trace of folding in this extensive range. It 
would rather seem as though the whole tableland plunged in a sir^le 
mighty flexure down to the gabbro of Lyngen^ and in fact Pettersen 
published some years ago a section taken a little more to the north, 
across the Lyngen fjord, in which the whole mass of stratified formations 
is represented as bending in an arc beneath this range of gabbro. At that 
time Pettersen even thought that the gabbro was intercalated with the 
stratified formations, and shared with them a common dip. I have not 
myself visited the mountains in the immediate neighbourhood of the 
gabbro range \ 

At the foot of the Ruostafjeld, mentioned above, a long valley opens — 
the Dividal — which comes from the south-south-east ; its right slope may 
be regarded as the continuation of the section just discussed. It is after 
this valley that Pettersen has named the lowest member of the high moun- 
tain series. Ascending the valley we met the Dividal stage, represented 
by shales which resemble in every particular the Cambrian shales of Ginee 
in Bohemia. We found no fossils, but typical plaited slabs such as occur 
in the 'primordial' deposits of Sweden. We followed the Dividal for 
a distance of 36 kilometers, always with mountains formed of flat-lying 
beds on both sides, and then turned to the west into the Skakterdal, where 
it joins the Dividal, and ascended it to the edge of the table-land, that is, to 
the great rampart. Here we perceived that to the south-east the Archaean 
foundation, and with it the whole of the superimposed series, gradually rises 
towards Sweden. At the bottom of the Dividal we saw the lowest members 
of this series ; further to the south Pettersen has even found the Archaean 
rocks exposed in the bed of the Dividal itself. From here (Hiitte Frihedsli, 
187 meters) we were obliged to ascend through the Skakterdal to a height 
of 724 meters in order to reach the junction with the overlying beds. 

Our way led us over broad rocky floors, polished and strewn with 
blocks which looked as though they had only yesterday been abandoned 
by the glacier, until finally the table-land terminated in two great cubical 
mountain masses. Store Jerta on the right, Namna on the left, which seem 
to guard, like towers, the mouth of the pass. Through this we step out 
into a moraine region dotted over with innumerable pools of water, the 
Tjoalma Vagge (lat. 68* 40' N.). 

Looking backwards from this desolate table-land we now see that the 

^ Petienen, Profil fra Rigsgrsendsen over Lyngen til KyalO; Forli. Vid. SeUk. 
Christiania, 1868, pp. 155-158, pi. 



CH.n] TJOALMA VAGGE 61 

great rampart is here divided iato a series of table mountains : JqIos 
Varre, Store Jerta, Nanms, Bumansberg, Kalbir Varre, and otheis. From 
the anterior slope of those nearest us, the Store Jerta, Namna, and Bumans- 
beig, there fidyanees with a particularly steep descent an intercalated bed 
of solid rock, which enables us even from a distance to perceive the original 
continuity of all these several bastions. 

Thos the great rampart is here broken up into batons ; and it is easy 
to see that it was through the gape between them, and in particular 
tfaxoogh that between Store Jerta and Namna, which leads down to the 
Skakterdal, that the ice-cap which once rested on the table-land thrust 
forth its gladets towards the Atlantic Ocean. 

Oatdde the bastions, on the barren table-land, below the moraines of 



Fib. 7. Jjaalma Vaggi, Uorairu landicapt on Ihi botiHdary bttteam Saneat and Smdtn 

(Ut 68* W N.). 

In tha foregTound the graund-down racAu tmiiiiotmia of the red Bwediah gTiLiiit«, 

in the diitancc the mperpoeed table moontain. 

Tjoalma Vagge, moutonn^ed bosses project which belong to the Archaean 
foundation. This is formed of a bright-red granitic rock, of distinctly 
banded stmcture; it strikes to the north, and dips at about 80' to the 
eaat. It is the source of those red blocks which have been carried out to 
the west into the fjord ; these will be discussed later. 

The ancient rocks of the Lofoten and Westeraalen, particnlarly the 
granite, gneiss and ancient mica schists, frequently accompanied by bosses 
of gsbbro, are continued in the same direction along tha Atlantic coast. 
They occur in the island of Amd, form that part of the peninsula of 
Heiland between the Rejsen and Kvanangen fjord which is tamed towards 
the sea^, as well as the whole northern part of the great peninsula of 
Beig, which follows next, then wholly or in great part the islands of 

' Petteiaen, Kvanangen, TronuO Hni. Aanh., IV, 1881, dgaciibei Bergt and 
MeilgjidihalvC, together with Spildem, Rode, and HoakO, oa onoe contmocniB rockj 
nnpart. 



62 THE OUTLINES OF THE ATLANTIC OCEAN [pabt in 

Stjemo, Seljand, Sor5, Evalo, as well as the Whole north-west part of 
the Porsanger peninsula, together with Hjelnfso, and we meet with theim 
on Magero extending up to the North cape. The islands, together with 
the most projecting parts of the mainland, thus form the continuation 
of the range of the Lofoten. The greater part of this range is 1,000 feet 
higher than the regions which succeed it towards the interior; it is broken 
across by the fjords. • 

Within this western range of gneiss lies the table-land, and the base 
of this, i.e. the lowest beds of the Dividal group, now also makes its 
appearance on the northern edge, as for example, on the east side of 
Evanangen, at Alten, on the south side of Lerbotn, in the Eomak Qord, 
and in the Bippe Qord on the west side of the Porsanger peninsula 
(lat. 70*" SO' N.). These exposures mark the boundary of the table-land 
against the western range of gneiss. In Lerbotn the Dividal beds are 
steeply upturned and the gneiss is not seen; in the Eomak fjord they 
lie horizontal, and the gneiss is also horizontal ; in the Bippe fjord they 
are intensely folded, while the older rocks are not folded, but exhibit 
fairly undisturbed bedding. With complete justice Pettersen concludes 
that the Bippe fjord is traversed by a great dislocation, and that the 
table-land lying towards the interior is thrown down against the gneiss 
ranged 

The Une of demarcation between the gneiae range and the table-land is 
thus a zone of dislocation, and this dislocation runs across those parts of 
the mainland which project into the sea and across a number of fjords. 
The table-land, here formed chiefly of schistose quartzite, extends from it 
over a wide area as far as the eastern rampart, so often mentioned above. 

While the gneiss zone as a rule has a height of S,000 feet, the table- 
land in its western part does not rise to more than about 2,000 feet. Its 
east border is little known. We left it in Tjoalma Vagge; Pettersen 
observed the base of the superposed beds on the upper Bejsenelv, not far 
from the Norwegian frontier; there the mountain mass of Beisduoddar 
Haldi, consisting of gabbro and serpentine, rises on the edge of the table- 
land to a height of 4,000 feet ^ 

According to Tellef Dahll's map the edge of the superposed beds now 
recedes a little, advancing again towards the south in the direction of 
Eautokeino. Leopold von Buch, after having crossed south of Alten 

^ Pettersen, De Notske Kyststrdgs Geologi, IV, PoraangerhalvGen ; Arch. Math. 
Naturrid., X, 1884, map, in particular p. 167. It is a pleasing fact that in the extreme 
north, in lat. 70^ 8(K N, where hardly a tree or even a bosh rejoices the eye, this 
obserrer has arrived by an unpregadiced contemplation of nature at the same views of 
the subsidence of great moiintain segments which have elsewhere, under much mora, 
fitvourable conditions, only been adopted after much discussion. 

* Pettersen, Ueber dasVorkommen des Serpentins und Olivinfels im N. Norwegen; 
N. Jahrb., 1876, pp. 618-622. 



OH.n] THE NORTH OF NORWAY 63 

horizontal beds of quarteite and mica schist, ascended this slope from the 
Ijohni-janre towards Eautokeino and there reached the granite ^. 

In the north we possess the observations of Tellef Dahll. According 
to these the boundary of the table-land runs from the neighbourhood of 
Kautokeino, along the Altenelv to the north, and on the river regular beds 
of graphite occur, intercalated with the stratified formations. Near the 
Jes Jaure, where serpentine crops out, the boundary turns to the north- 
east; it reaches the end of the Porsanger fjord and thence runs to the 
east. Raste Eaisse, south of the Lakset fjord, 8,000 feet high, and 
presenting beds of sandstone 2,000 feet thick, lies on this boundary, 
which reaches the western end of the Varanger fjord; the south side 
of this Qord belongs to the Archaean region, the north side to the 
table-land ^ 

In the north of the peninsula we may thus distinguish three chief 
elements as we proceed from west to east : the great gneiss zone of the 
Atlantic coast, which we call the zone of the Lofoten, then the table-land, 
which in spite of its poverty in organic remains has been proved to belong 
to the Cambrian system, in part even to the lower Silurian, and finally 
the Archaean land on the east. The western boundary of the table-land 
is a fracture, the eastern corresponds to normal superposition. 

Towards the south the line of the Lofoten abandons the direction of 
the coast, but the investigations of Pettersen show that the masses of 
gneiss and granite, of which the Lofoten islands consist, are also present 
on the east side of the Vest fjord, and this great trough has been faulted 
down into the rocks of the Lofoten, the so-called * coast-granite.' Further 
to the east, on the mainland, we find the same stratified rocks as we have 
just encountered about lake Tome& and the Bals fjord, but here they are 
folded. Gneiss and a long band of granite (the Ejolen granite) crop out 
in the anticlinals. The folds strike nearly parallel to the coast; they 
form the mountain of Sulitjelma ; they are intersected transversely by the 
QordSy as for instance by the Salten fjord. Here therefore a folded region 
lies between the range of the Lofoten and the taUe-land \ 

Further to the south the folded region forms the Atlantic coast almost 
as far down as the mouth of Hardanger fjord (lat. 60"" N.). It here con- 
sists, apart from intrusions of eruptive rocks, partly of beds of the same 

^ K Ton Bach, Reise durch Norwegen und Lappland, II, Berlin, 1810 ; Gesanunelte 
Schriften, edited by Ewald, Roth und Eck, 1870, II, pp. 449-453 and the section pi. v. 

' Tellef Dahll, Dm Finmarken's Geologi ; Forh. Yid. Selsk. Christiania, 1868, pp. 213- 
222, pL iv. T. Dahll rightly recognizes here the equivalents of the Cambrian alum 
shales ; he regards the overlying beds as Devonian, the graphite beds as altered Coal- 
meararea of the Carboniferous period, and the brown sandstones, superimposed on these, 
as Permian ; no fossils have so far been found. 

' Petteraen, YestQorden og Salten ; Arch. Math. Naturvid., 1886, XI, pp. 377-492, 
inap. 



64 THE OUTLINES OF THE ATLANTIC OCEAN [partui 

age as those of the table-land (as is proved by the Silurian fossils found in 
many places), and partly of the rocks which underlie these beds. 

The two transverse sections so far published, as for instance that of 
Trondhjem by Meraker', scarcely enable us as yet to form any definite 
opinion as to the direction in which tangential force has acted : but the 
laborious surveys of this extensive and sparsely populated mountain region 
accomplished by Ejerulf and his fellow workers show clearly that the 
folds strike down from the north towards the south-west into the imme- 
diate neighbourhood of that part of the country which projects furthest to 
the west, the promontory of Stat (lat. 62° 10' N.), without however follow- 
ing exactly the course of the coast. Towards the south, indeed, according to 
Ejerulf's map, the folds appear to turn more and more to the west, so that 
the Vigten islands, and next, with a more westerly trend, the great islands 
to the south outside the Trondhjem fjords indicate the strike of the folds 
as they pass out to sea, and the more south-westerly course of the coast 
towards Stat appears equally to correspond with a swerving of the folds. 
Just above Stat promontory, however, an unexpected deflexion occurs. 
The principal band of gneissose granite, which comes down from the 
north-north-east in the direction of the Vartdals fjord, runs to the Vanelvs 
fjord, then bends up towards the north-west, and so reaches with this 
direction Stat promontory and the island of Sando ^. This is the begin- 
ning of a change in direction. To the south of Stat the strike is due west, 
and correspondingly the fjords also here assume an east to west direction ; 
till finally, still further to the south beyond the Sogne fjord, the folds 
which appear on the coast strike into the interior, describing a semicircle 
around the neighbourhood of Bergen, and to the south of Bergen they 
reach the sea for the second time. 

In its southern part, this semicircular curvature has been described in 
great detail by H. Reusch ^. It is not a curvature with torsion like the 

> T. Kjerulf, Merakerprofilet, E. Norsk. Vid. SeUk. Skr., 1882, pp. 68-140, 7 pi. ; 
F. Svenonius, N&gra profiler inom mellersta SkandinavienB Skiiferomr&de, Geol. FOreu. 
Stockh. FOrh., 1885, VII, pp. 631-653, pi. xvii. FobsUb from these regions have been 
described by BrOgger, Om Trondhjemfeldets midlere Afdeling mellem Guldalm og 
Meldalen ; Forh. Vid. Selsk. Christiania, 1877, no. 2, pp. 1-28, pi. and maps. 

^ Hans H. Reusch, GrandQeldet i sOndre S6ndm0r og en Del af NordQord ; Forh. Vid. 
Selsk. Christiania, 1878, no. 2, pp. 1-18, map. 

' Hans H. Reusch, Silurfossiler og pressede konglomerater i Bergensskifrene, 
Universitets Program for 1 Halvaar 1883, 8vo, Christiania, 1882; by the same. Die 
fossilienftlhrenden krystallinischen Schiefer Ton Bergen in Norwegen, translated into 
German by Baldauf, Leipzig, 1883. These investigations have furnished excellent 
examples of the extreme alteration of the rock by pressuie and of the preservation of 
fossils in such rocks; they exactly correspond with observations made in the Alps. 
I was, however, much struck by the fact that in the flat-bedded parts of Norway also, 
particularly in the north, mica schist and hornblende schist occur, as has been already 
mentioned by many writers, and that in the Tromsd district for example, as Pettersen 



CH.n] 



GUNT LINES 



65 



great Bonmanian arc on the lower Danube (I, p. 482), and indeed I know 
of no other example of the same kind. To the south of it, on the Har- 
danger Qord the folds terminate, and the horizontally stratified belt 
extends to Stavanger. 

The way in which the folded ranges strike out to sea, in particular the 
bending round near Stat, the westerly strike of the folds south of Stat, and 
the lie of the Bergen arc, show that the continuation of these mountains 
lies to west and south-west beneath the sea. The direction of the struc- 
tural features in west Scandinavia points towards the Shetland and Orkney 
islands, but the discussion of the remarkable correspondence which exists 
between Scotland and the mountains of Norway may be postponed to 
a later occasion. 

3. Glint lines. On both sides of the Atlantic Ocean, to the north-west 
and to the north-east, there lies a vast region of Archaean rocks, from 
which the sedimentaiy covering, except for a few isolated patches, has 
been removed by erosion* In each of these regions there lies a shallow 
sea, on one side Hudson bay, on the other the Baltic. Each region is 
surronnded as with a rampart by the basset edges of the sedimentary 
formations, elsewhere swept away; these sediments everywhere belong 
to the older divisions of the Palaeozoic period, at latest to the Devonian, 
and round both shields they lie flatly bedded. In Bussia the feature 
formed by these basset edges is known as the gli/nt, and we shall in future 
employ this term for long lines of escarpment formed of flat-lying beds, 
when iJiey are due, not to f ractxire, but to denudation \ 

The glint of each of the two shields is marked by a long series of fresh- 
water lakes, which lie in its course. 

The glint lakes around the Canadian shield are: Ontario, Georgian 
bay, lakes Superior, Winnipeg, Athabasca, Great Slave lake, Martyr lake, 
Great Bear lake ; the glint line then runs out into Coronation gulf, which 
is situated like a glint lake* 

The glint line of the Baltic shield runs close to the east coast of Sweden 

nfltkHj observes, the lowest of these beds show the least alteration. Indeed, I have seen 
at the sonthemmost end of the Bals Qord, near the place where the great flexure (Fig. 6, 
pp. 58, 59) reaches the sea, manj signs of striation and gliding on the bedding planes, 
and have found in some places, as at the bottom of the Dividal itself, partings in which 
the layers are plicated into themselves, as it were, although the other beds have 
remained horizontal; we may thus conclude that during the formation of so great 
a flexure there occur such tensions and interior movements of the beds one over the 
other that great alteration of the rocks may be produced without any real folding ; but 
there are vast regions where such flexures do not occur, and yet through nearly the 
whole length of the peninsula crystalline schists are known at a higher level than 
deposits which are undoubtedly of Cambrian age. 

1 The outcrops described by Zittel in the Libyan desert are glint lines, produced by 
ibnnon ; the declivities at the south foot of the Uinta, which were produced by flexures, 
are not so. 



mujkM n 



F 



66 THE OUTLINES OP THE ATLANTIC OCEAN [pabtih 

not far from Ealmar, then through the sea north of Qothhind, north of 
Dago, throng the gulf of Finland, which also has the position of a glint 
lake, through lakes Ladoga and Onega, into the gulf of Archangel and bo 
to the Arctic Ocean, 

A closer inspection of the structure of the Scandinavian peninsula 
shows that, in the north at least, north of Jemtland or of lat. 64'' N., the 
eastern face of the table-land with its slightly inclined bedding, its bastions 
and ancient glacier portals, forms part of the Baltic glint. Just as, many 
years ago, Richardson and Isbister in north Canada remarked with as- 
tonishment the presence of so many lakes on the Palaeozoic boundary line, 
so, many years ago Tomebohm observed here that most of the great lakea 
of Lapland are intersected by this line of escarpment ^. Indeed, all these 
great lakes, such as Homafvan, Saggat Trask near Evikkjokk, Lule& 
Jaur, Paitas Jaur, Tomei Trask, Alte Vand and Buosta Vand, must be 
regarded as a series of glint lakes. The glint rises higher here than 
elsewhere, and the transverse position of the lakes is more strikingly 
shown than in any other region of similar character. The majority of 
these lakes find an outlet for their waters into the gulf of Bothnia, the 
remainder into the Atlantic Ocean, but this is a point which we need not 
consider at present. 

The glint line is continued through the north of Finmark to the end 
of Varanger fjord, and through this to the Arctic Ocean. 

From the foregoing, we perceive that there are not only glint lakes, 
but also gUnt gulfs of the sea. Such are : the gulf of Finland, the gulf 
of Archangel, Varanger fjord and Coronation gulf. Of these four gulfs, 
three belong to the Arctic Ocean. 

4. The taUe4and of SpUzbergen. The Russian chain of lakes reaches 
the Ocean in the gulf of Onega and the White sea, immediately to the 
west of the peninsula of Kanin, the relations of which with tixe Ural 
range have already been discussed. 

We have already learnt that Nova 24embla is a mountain chain folded 
to the west and north-west, which joins the Ural near Eonstantinov- 
Eamen in lat. GS"* 29' N. (I, p. 504), the shar of Jugor, the straits of Kara, 
and the Matotshkin shar are transverse furrows, and the Kara sea is 
a * back-sea' (Biickmeer). The Timan range, a divergent fore-fold or 
parma of the Ural, unites in syntaxis on the Sula with a little fragment 
of an arc, which strikes from the north-east of the Tshesskaja bay across 
the peninsula of Eanin towards Eanin Noss (I, p. 505) K 

^ TOraebohm, Geognostik Bchwediacher Hochgebirge, p. 49. 

* A. Wichmann has expreised doubt as to Hofer's conception of the stractore of Nova 
Zambia ; but since Wichmann himself regards the island as the direct oontiniiation of 
the island of Waigatsch and of the Pae-chol, this settles the chief question. I have 
recently rediscussed the matter with Herr Hofer, and am the more convinced that his 



CH. n] DEPTH OP THE SEA ABOUT SPITZBERGEN 67 

We have thus good reason to suppose that the valley of the Petchora, 
the floor of which is formed in great part of fiat-lying Jurassic beds, 
stretches across the flat island of Eolguev to be continued beneath the 
south-east part of the Barents sea: this gives all the more interest to 
a knowledge of Spitzbergen, Franz-Josef Land, and Bear island. 

Those who, like Nordenskjold, Mohn and Nathorst, are best acquainted 
with these regions, have conjectured that a connexion once existed between 
Spitzbergen and the north of Norway; and Pettersen, who has brought 
together all the observations relating to this subject, distinguishes the 
table-land, or the group of great islands which may once have occupied 
the site of the existing Barents sea, as the 'Arctis^.' 

It is to the west of the Lofoten islands, on the Vesteraals-Eggen, that 
the sea first sinks to considerable depths along the Norwegian coast, as 
has been shown with great clearness by Mohn. The steep submarine slope 
runs thence to the north, towards the west coast of Spitzbergen. Barents 
sea lies above it, outside and to the east of the great depths of the Oreen- 
land sea, which sinks between Norway and the recent volcanic island of 
Jan Mayen to below 2,000 fathoms, and between Spitzbergen and Greenland 
to below 2,600 fathoms. Barents sea is thus of trifiing depth ; a furrow 
about 200 to SOO fathoms deep runs up between Norway and Bear island, 
and this is its deepest part ; the hundred fathom line, however, runs from 
the Murman coast towards Nova Zembla, then from the north-west part 
of this island to the west and includes Bear island; so that the islands 
mentioned above, extending from Spitzbergen to Franz-Josef Land and to 
Bear island^ form the highest parts of a single continuous plateau. 

The results published by the Norwegian North-sea expedition show 
besides, that lying on the submarine slope which descends to the grc^t 
depths from lat. GS"" S6' to 68^ 2V N., and in isolated localities even up to 
lat. 78"" N., there are some residual fragments of Cretaceous, although 
deposits of this system are not known at present on the neighbouring land. 
These consist of pieces of chalk and flint; in lat. 65'' 48' N., long, r 29' E. 
a fragment of a Belemnite was obtained from a depth of 194 fathoms. 
These fragments either prove the existence of a submarine Cretaceous 
belt, or excessive denudation of the surrounding land ^. 

Let us now turn our attention first to the archipelago of Spitzbergen. 
Through the great kindness of Professor Nathorst, of Stockholm, I am 
able to make use of his comprehensive, and, as yet, unpublished observa- 

▼iewB are conect since they fully accord with what we know of the structure of the 
north Ural; A. Wichmann, Znr Geologic von Nowaja-Sem^ja, Zeitschr. deutsch. geol. 
Gea^ 1886, XXXVUI, pp. 516-550. 

^ K. Pettersen, Arktis, II ; Arch. Math. Naturvid. Eristiania, 1882, pp. 465-489. 

* H. Mohn, Dybde-Kart over Nordhavet ; Norske Nordhavs-Expedition, 1887, rVIII, 

F 2 



68 THE OUTLINES OP THE ATLANTIC OCEAN [pari in 

tions on the stratified Beries and the structure of Spitzbergen. The 
following deacriptioa is derived from commu&ications he has kindly made 



Fio. 8. SpittbtrgiH. 

me hy letter, and afibrds a more complete picture than the valuable works 
of Nordenskjold, Hiif er, Draache, and others '. 

* In particular, A. E. Noidenakjold, Spetsbei^n'B Geolo^e, K. Svenikft Tet. Akid. 
Handl Stockholm, 1867, VI, with geological map, alio an English tronalation in the 



cH.n] SEQUENCE IN SPITZBEBGEN 69 

Aecording to Nathorst the stratified sequence of Spibsbergen is as 
follows :— 

The Tertiary deposits oonsiBt, in the first place, of about 500 feet of 
lacostrine sediments with coal and land plants ; beneath these lie 2,500 feet 
of marine sandstone, clay, and other beds, with marine moUnsca, and 
finally another 100 feet of coal and plant-bearing beds. 

Beneath these Tertiary beds a gap occurs with unconformity, then 
comes the Jurassic series, first, marine beds with Leda, Nucula and others, 
and then lacustrine beds with plants and fresh-water moUusca (once erro« 
neoosly believed to be Cretaceous); next, marine beds with Ammonites; 
again, limnic beds with coal and plants ; below these again, sediments which 
ll^haps marine. The JtuJic teste conf onaablyTmarine sedimente. 
also completely conformable among themselves, of the Trias, Permian, and 
Permo-Carbonif erous ; the last of these rests also conformably on the Ursa 
stage, which contains plant remains, but in one locality it also exhibits an 
intercalation of marine sediments. 

The Ursa stage rests unconformably on the Devonian series of liefde 
bay, which contains Cephalaspis, Scaphaspis, and plant remains of the Old 
Bed sandstone : in this series also a marine bed is probably intercalated. 

The Devonian rests unconformably on the much older Hecla Hook 
system, which has not as yet afforded fossils; it consists of quartzite, 
phyllites, limestone, and dolomite, and close to it there occur granite, 
gneiss, and hornblende schists, which are said to represent the Archaean 
foundation, though Nathorst does not consider this definitely established. 
The Hecla Hook rocks which I saw in Stockholm resemble the phyllites of 
the eastern Alps. 

Of the more recent eruptive rocks only diabase occurs ; it traverses the 
beds as far as the Trias, in places even to the Jurassic. The largest masses 
of diabase occur in a zone running from south to north, which extends 
from the Thousand islands along the Star fjord far into the straits of 
Hinlopen. 

The dislocations resulting from tangential stresses are confined to the 
fundamental rocks, and the Hecla Hook beds. These are, as a rule, steeply 
upturned, and form the sharp jagged mountains to which the highest peaks 
of Spitzbei^n, such as the Homsundstind, belong. All the other deposits, 
however, from the Devonian upwards, retain, except on the west side of 
Spitsbergen, a horizontal or only gently inclined position, save for some 
local and insignificant disturbances. The whole of the country formed by 
them is table-land. 

GeoL Hag., 1872 ; H. HGfer, Giaf Wilczek's Nordpol&hrt im Jahre 1872, 1. : Beitrftge zar 
Geographic SM-Spitzbergens, Peterm. Mittheii., 1874, pp. 220-223 ; R. yon Diasche, 
Geologische Beobachtungen aof einer Reise nach den WestktlBten Spitibeigens im 
SoBuner 1878, Yerh. k. k. geol. Beicha, 1873, pp. 260-868. 



70 THE OOTLINES OP THE ATLANTIC OCEAU [pAHrm 

Along the west side of Spitzberg«n there mna a long hotst which, 
towards the east, is separated from the ishmd by a great line of fracture. 
On this line the sedimentary beds are steeply aptomed or even inverted ; 
according to Nathorst the feature mnst be regarded as a faolt with down- 
throw to the eaet accompanied by apward flexure. The throw of the 
dislocation most amount to several thousand feet, since on the Ice sound, 
where the flexure is most widely extended, even the Tertiary beds are 
afl'ected ; the same is the case in King's bay. In Bel sound the upturning 
appears to terminate with the Trias. 

The greater part of Spitzbergen is thus a table-land, and the same 
would seem to be true of Barents Land, Edge island, and Hope island, since 
all the beds of these islands appear to lie horizontal. 

Drainriiut in Another great fault runs down through 

th« Wijde bay towards Klaas Billen bay. 

^ t trough ^pyg disturbance is, however, certainty 

very ancient, even older than the Ursa 

stage, since, west of Klaas Billen bay, beds 

J of this stage tie flat on the Devonian, 

which is somewhat steeply upturned 

against the fracture, and to the east of 

this bay they rest immediately on the 

fondamental rocks still retaining their 

flat bedding. In Liefde bay, Elaas Billen 

bay, and Dickson bay, Nathorst foond the 

Devonian everywhere horizontal, and the 

whole Devonian table-land appeared to 

him to be let down in an old pre-Carboniferous fault-trough. 

It thus f^pears that faults are present of veiy different age ; some are 
pre-Carboniferoua, while others are post-Tertiary, and have produced an 
upward flexure of the Tertiary beds. 

The throw of another fault, situated in Bel sound, along the north 
coast of Van Kenlen's bay, must, according to Nathorst, amount to at least 
2,000 feet, for beds of Carboniferous limestone are exposed both in the 
mountain peaks and on the seashore below. 

Other faults of trifling importance also occur, some of which coincide 
with the lie of the fjords. 

These mfdn lines of the structure appear in the tittle map, Pig: 8, 
reduced from a sketch which Professor Nathorst was kind enough to send 
me. The ancient Hecla Hook rocks form almost the entire west ooaafc, 
probably also Prince Charles promontory ; the Archaean foundation makes 
its appearance in the north-west." Between Liefde bay and Wijde bay the 
Devonian is exposed, bounded on the east by the fracture, and on the whole 
north coast the ancient rocks of Hecla Hook and the Archaean are visible. 



CH^n] TABULAR FRACTURES OP SPITZBERGEN 71 

The centre of the country and the whole eastern part from the South cape 
to beyond Edge island is occupied by later sediments, and by the Mesosoic 
exposures of diabasa 

It is extrranely likely that the eastern islands are the continuation of 
the Mesoosoic table-land in which Stor fjord lies. 

As to King Charles' La/nd Nathorst observes that the only fossil so far 
known is a fragment of fossilized wood, described by Schroter as La/ria 
Joknseni, and believed to be Miocene ; that Tertiary beds occur here, is 
certainly possible, but in Spitzbergen a conifer is found which probably 
belongs to Larix, and this occurs in the upper Jurassic* 

In FruTiz^oaf^ Land the bold sledge journey of Payer has made us 
acquainted with horizontal sheets of eruptive basalt, and the more recent 
^vestigations of Leigh Smith have still further extended our knowledge. 
In Eira harbour (lat. 80** 10" N. Franz-Josef Land, S.W.), Qrant, who accom- 
panied Leigh Smith, found that the cliff, 1,040 feet in height, is formed at 
its base of Oxford clay with Belemnites. Above this follow beds with 
eonif eroos wood and cones, and other plant remains ; they are believed to 
be Cretaceous. The whole is covered with a sheet of basaltic lava \ 

Let us now turn our attention from Spitzbergen to the south. 

Bear idand consists of horizontal beds of Permo-Carboniferous age, 
and the Ursa series. Nordenskjold met here with the Heda Hook rocks ; 
according to Nathorst it is uncertain whether these are exposed by erosion 
or by a prolongation of the great line of fracture which runs along the 
west side of Spitzbergen. The island is nothing but a fragment of the 
Spitzbergen table-land ; Mohn has given a vivid account of the unceasing 
attack of the breakers on these horizontal beds, the undermining of the 
difb, the formation of great clefts, and the slipping down of mighty masses 
of limestone ; such fallen masses, standing in isolated pillars or towers at 
the end of jQat tongues of land, often remain to bear a long-enduring 
testimony to the destructive violence of the waves ^. 

Thus we find repeated on this island that castellated form of the cliffs 
which has been so often described as characterizing the Palaeozoic limestone 
beds, likewise horizontal, of the North American archipelago, e. g. in the 
Pany islands and Lancaster sound. There we found arms of the sea 
numing up in many ramifications between the islands of the archipelago, 
like a submerged system of river courses. We can indeed picture to 
ourselves how, with a rising strand-line, the sea will penetrate into the 
valleys of a horizontally stratified land, how it will widen them by under- 

' C. & Markham, The Voyage of the ' Eiia,* and Leigh Smith's Arctic DiBCOveries in 
1880; Ptoc. Geogr. Soc, London, new series, III, 1881, p. 135. According to Camithera 
the conifer is a tme pine. Etheridge mentions still older rocks beneath the Jurassic ; 
torn, cat, p. 147. 

' Nor^e NordhaTB Expedition, Y, 1882 ; H. Mohn, Geogntfi og Naturhistorie, p. 82. 



72 THE OUTLINES OP THE ATLANTIC OCEAN [paktiii 

mining their banks, and so produce outlines which, as soon as the secondary 
watersheds have been crossed or destroyed, may come at last to resemble 
and not remotely those of the Parry archipelago. In the same way Bear 
island is regarded by most observers simply as an outlier of the great 
table-land, separated by erosion. Thus the sea completes the work which 
the rivers had begun, and in this way are accomplished those extensive 
abrasions to which the stratified succession of so many lands bears 
witness. 

5. Oreenland, The fundamental features of the stratified succession in 
Spitzbergen are repeated in Greenland, but only on the east coast and 
a part of the south-west. This correspondence has often been remarked 
upon by competent observers. Here as there, red sandstone rests hori« 
zontally and unconformably on ancient folded rocks, and the folding is 
thus older than the sandstone. In Greenland the red sandstone has 
not as yet furnished any organic remains, but it may doubtless be corre- 
lated with the liefde bay series of Spitzbergen. The Mesozoic and 
Tertiary deposits also lie flat, as in Spitzbergen. This is the constitution 
of the great peninsula in the east, so far as the inland ice and the in- 
hospitable nature of the coast have permitted us to become acquainted 
with it; in the west the characters of an abraded table-land are con- 
tinued far to the north, probably about as far as the Humboldt glacier, 
where, as we have ahready seen, the upturned Silurian strata extend from 
Norman Lockyer island (lat. 79*" 25' N.) in Scoresby bay across Smith sound 
with a north-easterly strike (II, p. 42). 

Greenland, as far north as Humboldt glacier, is probably an ancient 
table-land, yet it differs in some respects from the Canadian and Baltic 
table-lands, for in these two regions we meet with horizontal Silurian beds 
of marine origin, while in Greenland as in Spitzbergen the series of flat- 
lying beds does not begin, according to the present state of our knowledge,, 
before the red sandstone. 

Payer and Copeland have explored the east coast between lat. 73"* 
and 76* N. ; the facts brought together by Hochstetter and his fellow 
workers show that the greater part of the coast of the mainland, together 
with most of Clavering island and Euhn island and the north of Shannon 
island, consists of Archaean rocks. Ancient quartzite, clay slates, and lime- 
stone, which are assigned to the Hecla Hook series of Spitzbergen, are 
exposed on the north coast of Franz-Josef fjord In False bay (Clavering 
straits, west of Sabine island) Payer made the remarkable discovery of 
Bhaetic fossil, Rhynchxmella fisdcodata, the only representative of this 
series found as yet in the Arctic regions. On Euhn island there are 
Jurassic deposits similar to those of Spitzbergen, and similarly associated 
with coal and plant-bearing beds. Basalts and plant-bearing sediments 
of middle Tertiary age crop out for a great distance along the coast from 



can] GREENLAND 73 

the little Bontekoe islands in front of Franz- Josef fjord, northwards up to 
the middle of Shannon island, so that all the land which runs farthest 
oat to sea — such as cape Broer Buys, cape Borlase Warren, Sabine and 
Pendulum island, and cape FhiUp Broke on Shannon island — belongs to the 
basalt region. Hochstetter foreland also is middle Tertiary : it has furnished 
remains of Tertiary marine mollusca similar to those of Spitzbergen. I may 
observe that these are the only two regions in the Arctic Ocean which 
have afibrded marine Tertiary fossils. The basalts of Iceland, the Faeroes, 
and even Ireland (so far to the south) and the west coast of Greenland are 
everywhere accompanied by Tertiary beds which contain no other fossils 
than terrestrial plants ^ (I, p. 287). 

The west of Greenland, from cape Farewell to lat. Gl"" N., consists 
mainly of granite, beside which gneiss appears only in isolated localities — 
aeoording to Laube on the precipitous cape Whitsuntide on the east coast, 
and on Sermersoak in the south-west. On the east coast the granite 
frequently contains hornblende; on the west, according to the recent 
investigations of Steenstrup and Eomerup, it is traversed north of Igaliko 
igord by a mass of syenite, and north of Julianehaab, at the mouth of 
the Tmmgdliarfik, by a mass of sodalite syenite, which is cut right through 
the middle by the fjord. A great mass of red sandstone, accompanied by 
porphyry, separates, north of the sodalite-syenite, the Sermilik from the 
Tnnugdliarfik fjord, and extends across the latter to the north end of 
the Igaliko Qord. So far the red sandstone, as we have already mentioned, 
has not afforded any organic remains ^. 

In the interval between lat. 62° 15' and 64** 15' N. Eomerup en- 
countered only Archaean rocks, chiefly grey gneiss. Its strike is generally 
directed to the north-east \ 

The land much further to the north also, between lat. 66** 15' and 
68* 15' N., was found by the same observer to be formed of different 

* F. Y. Hochstetter, Geologie Ost-Grdnlands zwischen dem 73. und 76." n. Br. : a. Allge- 
meine Uebersicht, bearbeitet von Franz Toula, h, Specielle Darstellung, bearbeitet von 
Oscar Lenz ; Mesozoische Verateinemngen von der Euhinsel, von Franz Toula, in Zweite 
Dentache Kordpolfahrt, II, 1872, pp. 471-511, and Hochstetter, Geologische Eartenskizze 
von Ost-GrrGnland nach den Beobachtungen und Sammlungen von Payer und Copeland, 
op. cit., pL i. All the earlier geological literature of Greenland is given in Rupert 
Jones' valuable manual of the Natural History, Geology, and Physics of Greenland, 8vo, 
London, 1875. 

' Gurtav Laube, Geologische Beobachtungen gesammelt w&hrend der Reise des 
'Hansa' und gelegentlich des Aufenthaltes in Stld-Gr6nland, Sitz. k. Ak. Wiss. Wien, 
1873, LXVIII, pp. 17-109, and geological map ; K. J. V. Steenstrup og A. Eomerup, 
Beretning om Ezpeditionen til Julianehaab*s Distrikt i 1876, Meddels. om Grdnland II, 
Sjdbnhavn 1881, pp. 1-26 ; Steenstrup, Bemserkninger til et geognostik Oversigtskaart 
over en Del af Julianenhaab's Distrikt, op. cit., pp. 27-41, and geological map. 

' A. Komerap, Geologiske lagttagelser fra Vestkysten af Gr6nland (lat. 62'' 15 - 
64'' 15^ N.) ; Meddels. om GrOnland, 1, 1879, pp. 77-139, geological map B. 



74 THE OUTLINES OP THE ATLANTIC OCEAN [pabt m 

varieties of gneiss, and this great mass of gneiss, lying in folds striking 
to the north-east, appears to form by far the larger part of the west coast ^. 

Between lat. 70"" and ^T 39' N. and beyond, the same Archaean rocks 
crop ont at the edge of the inland ice, but in front of them, towards the 
sea, Ue great basaltic regions, such as we have already studied on the east 
coast from the mouth of Franz-Josef fjord up to Shannon island I may 
cite as examples taken from the latest surveys of Steenstmp, Disko island 
(on the south and south-west coast of which the Archaean foundation is 
visible), then Hare island, the peninsula of Nugsuak, Ubakjendt island, the 
peninsulas of Svartenbuk and Ingnerite. On this side as in the east, the 
Archaean region of the interior is bounded by a basaltic belt, and it is 
beneath these mighty sheets of lava that those rich Cretaceous and middle 
Tertiary floras are preserved which have furnished the most important 
material for the investigations made by Oswald Heer on the Arctic floras 
of the past \ 

A fact of great importance in its bearing on questions to be discussed 
later is the discovery made by Steenstrup of the presence of marine Cre- 
taceous fossils within this series of beds which, except for some local 
disturbances, are always horizontal The succession, according to the 
present state of our knowledge, is as follows: — (1) Kome beds, resting 
directly on gneiss, and containing a terrestrial flora of lower Cretaceous 
age; (2) Atane beds, with a terrestrial flora of Cenomanian age, corre- 
sponding to that of the Quader sandstone of Central Europe, and containing 
several species of the flora of the Dakota stage in America; (3) Patoat 
beds, containing a terrestrial flora of Senonian age, and also, up to 1,200 
feet above the existing sea-level, Cretaceous marine fossils. Loriol has 
established their affinity with the Fort Pierre and Fox Hill beds of 
Nebraska ^ Above these lie the Tertiary plant-bearing beds, and the 
great basaltic sheets which reach a height of more than 5,000 feet above 
the existing sea. 

The Cretaceous marine beds are known in the south of the peninsula 
of Nugsuak, and also in the north up to about lat. 7ff* 45' N. Thus far 
therefore the Senonian sea had penetrated. We mentioned a patch of 
marine Cretaceous on the Mackenzie in lat. 65*" N. and the fossils of Patoot 
point clearly to a connexion with the Cretaceous sea of the North American 
prairies. , Although we shall scarcely be able to determine how far the 
Cretaceous sea once extended to the east, over the abraded Canadian 

^ A. Eornerap, Geologiske lagtiagelser fra Vestkysten af GrOnland (lat. W 55'- 
eS*" 15' N) ; op. cit., II, 1881, pp. 151-208, geological map, pi. vi. 

^ E. J. Y. SteenBtnip, Bidrag til Kjendskap til de geognostdske og geografiske Foihold 
i en Dal af Nord-GrOnland, Meddels. IV, 1888, pp. 178-242, geological map ; and by the 
flame, Om Forekomsten af Forsteninger i de EulfOrende Dannelser i Noid-Gr0nland, op. 
cit, V, 1883, pp. 45-77. 

' P. de Loriol, Om fossile Saltvandsdyr fra Noid-Grdnland ; torn, dt., pp. 203-213. 



CH.nJ THE CALEDONIAN MOUNTAINS 76 

taUe-Iand (I, p. S68), yet in the extreme north we recognize it in isolaied 
traces beneath the basaltic sheets. 

The earUer observations of Sutherland, who was akeady acquainted 
with the basalts from Disko np to Proven island in lat. IV W N«, show 
that the coast to the north of Proven island, as far as cape York in 
latb 76* N., consists, together with the bordering islands, almost exclusively 
of gneiss and granite. North of cape York up to cape AthoU, i.e. for 
a distance of about 80 to 40 knots, basaltic outflows, probably the same 
as those of Disko, again make their appearance ^. 

Gape Alexander at the mouth of Smith sound consists of basalt, and 
this again is followed to the north by gneiss ; the rest of the east coast 
of Smith sound towards the Humboldt glacier appears to be unknown '. 

Along this great ice stream or close to its further side, the folded ranges 
with Palaeozoic fossils begin. The Tertiary lignite beds also extend into 
this r^on, and were encountered by Feilden in Qrinnel Land, at Discovery 
harbour in lat 81* 46' N. ^ 

Great as is the variety of sediments on the west coast of Greenland, 
yet as far north as the Humboldt glacier the only beds which have so far 
afforded marine fossils are the Senonian intercalations of Patoot ; a striking 
eontrast to the east coast, where indeed the Cretaceous formation has never 
yet been ekicountered, neither has it been met with in Spitzbergen. 

6. The Caledonian Tnountaina. ^The Scottish 'Highlands, with the 
Hebrides and Donegal on the one .hand, and the Orkneys and Shetlands 
on the other, must be regarded — ^to use a technical phrase— ^U9 mere " out- 
liers " of the Scandinavian peninsula.' So writes Judd, one of the geologists 
best acquainted with the country ; and other distinguished investigator8,such 
as A. Geikie, hold the same view ^. This conception of the Scotch geolo- 
gists rests, it is true, primarily on the correspondence between the stratified 
series of the two regions. The Torridon sandstone, an arkose, or sandstone 
containing feldspar, which underlies the Cambrian, is correlated with the 
sparagmite of Norway. A. Geikie is inclined to think that the zone of 
Old Bed sandstone, which has been traced along the Moray firth through 
the east of Boss and Sutherland, then through Caithness to the Orkneys 
and into the southern part of the Shetland islands, may once possibly have 

' P. C. Sutherland, On the Geological and Glacial Phenomena of the Coaats of Davis' 
Strait and Baffin's Bay ; Quart. Journ. Geol. Soc, 1858, IX, p. 297. I have not ventured 
io quote the observations made north of cape Atholl, since they are rather indefinite and 
evidently made from the ship. 

* The little map by Feilden and de Ranee in Quart. Journ. Geol. Soc, 1878, XXXIV, 
pi. xxiv. 

* O. Heer, Notes on Fossil Plants discovered in Grinnell-Land by Capt. H. W. Feilden ; 
torn, dt, pp. 66-70. 

* J. W. Judd, Address io the Geological Section of the British Association at Aberdeen, 

1885. 



76 THE OUTLINES OP THE ATLANTIC OCEAN [paktih 

extended into the Sogne fjord and DaLs fjord, where similar red con* 
glomerates have been met with, although so far without organic remains K 
Judd points out the striking resemblance that exists between the Mesozoic 
deposits of Scania and those rare patches of the same age in Scotland, 
which have been preserved by great subsidences or beneath sheets of 
basalt. In both regions the series begin with Trias sediments (containing 
reptiles in Scotland), and above this there follows a series of alternating 
marine and fresh-water deposits, the latter containing the remains of 
terrestrial plants; in both regions the lowest members of the Cretaceous 
are absent, and the higher members are alone represented. At Ando, in 
the Lofoten islands, the Jurassic no doubt owes its preservation to subsi- 
dence, as it does on the Scotch coast. 

Our study of the Scandinavian peninsula has shown that it may be 
distinguished into several regions. The first is the Archaean zone of the 
Lofoten or the western zone of gneiss, which may be traced across the 
islands and peninsulas of the north, from Magero to Vest fjord. The 
second region is the table-land of the north, separated by dislocation 
from the western zone of gneiss and terminating on the east with hori- 
zontal stratification in the glint, which runs down from the Varanger fjord. 
In the Bippe fjord, lat. 70"^ 30^ N., the dislocation along the western gneiss 
zone is a fault with upward flexure of the down-thrown table-land ; at the 
bottom of the Bals fjord, on the other hand, in lat. 69" 10' N. all the table- 
land appears to dip westwards in a great flexure beneath the gneiss zone, 
or at least beneath the band of gabbro which lies in front of it. Further 
to the south the table-land itself in its western part appears to be thrown 
into the long, south-south-west striking folds of the great Norwegian ranges, 
while to the east it always remains flat and runs out to the glint. East of 
the glint lies the Archaean table-land of the gulf of Bothnia ; that is, the 
Baltic shield. 

These fundamental features in the structure of the Scandinavian 
peninsula enable us to perceive that there exists in the two regions not 
only a correspondence between the stratified succession as shown above^ 
but also a far-reaching correspondence in structure, and that the contours 
of the Atlantic from the North cape in lat. 71** 10' N. to at least as far as 
Donegal bay in lat. 54** 80' N. are formed by once continuous mountain 
ranges, constructed on a common plan. 

A. and J. Geikie have shown in a number of highly instructive descrip- 
tions that the structure of Scotland is clearly expressed in the configuration 
of the surface, as well as in the character of the landscape. It also reveals 
itself in the plainest manner in the course of the coast lines ^. 

' A. Geikie, Textbook of Geology, 8vo, London, 1885, p. 712. 

* Prom this long serieB I will only quote the latest work known to me : J. Geikie, The 
Physical Features of Scotland, Scott. Geogr. Mag., I, 1885, pp. 26-41. For the con- 



can] GNEISS OF THE HEBRIDES 77 

Scotland and the adjacent islands are represented in Fig. 10 on the 
same scale as that chosen for the sketch of Spitzbergen (II, p. 68, Fig. 8). 
We oheerve a continuous range of Archaean rocks in the north-west. The 
ancient gneiss is visible in the Hebrides, in the south-east of the islands of 
Gcdl and Tiree, in Baasay and Bona, and along the north-west coast of 
Scotland. In a precisely similar fashion the Archaean range of the 
Lofoten lies to the west and in front of the Scandinavian peninsula. 

The Minch lies on the gneiss region in the same way as the Vest fjord does 
in Norway ; some of the Tertiary volcanos of the Hebrides belong to the same 
r^on (I, p. 156, Fig. 19), and in particular the great basalt flows of Skye. 

The Archaean area is bounded on the east by a zone of dislocations 
of a peculiar nature; beginning on lake Eriboll it is continued through 
loch Assynt and the upper parts of loch Broom and loch Maree as far as 
loch Garron, and no doubt still further to the south-south-west, on the inner 
side of the gneissose islands of Coll and Tiree. Murchison, deceived by the 
extraordinary position of the beds in the north part of this zone, came to 
the conclusion many years ago that the gneiss is here normally superposed 
on fossiliferous Silurian beds, and he consequently asserted the existence 
of a more recent Silurian gneiss as opposed to the ancient gneiss of the 
Hebrides. Nicol challenged Murchison's views, and thought that the super- 
position of the gneiss was due to great dislocations. This interpretation was 
revived by Bonney, Hicks, and Callaway, and found its strongest support 
in the brilliant discoveries of Lapworth \ To bring a long controversy to 
a close, A. (}eikie,'as. the Director of the Geological Survey, commissioned 
Messrs. Peach and Home to make an exact survey of the district around loch 
EriboU and the extreme north-west of Scotland. As a result of this work, 
A. Greikie has now admitted, with the sincerity which distinguishes and 
honours the true man of science, the erroneousness of the older theory ^ 

The section from the Kyle of Durness across loch Eriboll published by 
Peach and Home reveals the following features : — 

timiatiofn to Irelaad, B. Harkness, On the Rocks of Portions of the Highlands of Scotland 
Soaih of the Caledonian Canal and on their Equivalents in the North of Ireland, Quart. 
Joom. Geol. Soc., 1861, XVII, pp. 256-271 ; E. Hull, The Physical Geology and Geography 
of Ireland, 8to, London, 1878, p. 20; C. Callaway, On the Granitic and Schistose Roclm 
of North Donegal, Quart. Joum. Geol. Soc, 1885, XLI, pp. 221-289 (Overfolding in 
Donegal, p. 238). 

* H. Hicks, On the Metamorphic and overlying Rocks in parts of Ross and Inverness 
Shires, Quart Joum. Geol. Soc., 1888, XXXIX, pp. 141-159 (Appendix : On the Litho- 
Icgical Characters of a Series of Scotch Rocks, &c., hy T. G. Bonney, torn, cit., pp. 
150-166), and map; C.Callaway, The Age of the newer Gneissic Rocks of the Northern 
ffigfalands, torn, cit, pp. 355-414 (Idthology in Appendix hy Bonney, pp. 414-420); 
C. lApworih, Close of the Highland Controversy, Geol. Mag., 1885, pp. 97-106 et passim. 

* A. Oeikie, The crystalline Rocks of the Scottish Highlands, Nature, Nov. 13, 1884, 
pp. 29-31, and B. N. P. Peach and J. Home, Report on the Geology of the North-west of 
Satheiland, torn, cit., pp. 31-35. The correspondence of the Lofoten islands and the 
Hebrides is emphasized by Bonney, Quart. Joum. Geol. Soc, 1870, XXVI, p. 623. 



78 THE Otn^INES OP THE ATLANTIC OCEAN [vast in 



Fia. 10. A GmtTTaSktUlii^Okt StructHn a/SaUandbiaedoa the Geologiea.lK»p at BootUai, 

bySlr A. Qeitie. 

a Arehasan ; ( Silurian and ancient sctiiita, togettier with included mBwnn of Omiit*; 

d Old B«d wndston*, sad ca CBrboniferoni, together with eruptive mMMO, whether Deronian 

>r Cuboniferoiu is still a matter of opituon ; (r Trias; ^Jurassic; Tertiary emptiTe rooks an 

ihownUocik. (The thruat- ' ..^-. .. -... .. . ....... 

la not completely iodioatei 



CBLn] OVERTHRUSTS ON LOCH EMBOLL 79 

In the neighbourhood of Durness, that is in the north-west, the 
Cambrian beds, probably 2,000 feet thick, folded and transversed by faults, 
rest upon the gneiss. Another range of gneiss advanoes towards the sea, 
and also bears in the east a patch of Cambrian. Then the country breaks 
op on the southern shore of loch Eriboll into numerous flakes of gneiss and 
Cambrian, the thrust-planes of which dip to south-east and east-south-east, 
and thus reveal a considerable tangential movement from this direction. 
Above these smaller flakes, and overthrust from this same direction, that 
is from the east-south-east, lies a mass of ancient gneiss, with a maximum 
thickness of 400 feet. This overlies, on a great plane of reversed faulting 
(thrust -plane of the Scotch geologists), all the various stages of the 
Cambrian one after the other, as they appear in succession in the smaller 
flakes. A patch of this overthrust mass of gneiss, separated from it by 
denndation, lies isolated on the Cambrian in the north-west of Ben 
AmaboU. 

East of Whitten head, at the mouth of loch Eriboll, there follows 
a second and yet vaster overthrusting of gneiss, which may be followed 
for many miles to the south-south-west — a second master-flake, so to 
speak, superposed on the firsi 

So far we have traced the overthrusts for a distance of 90 miles to 
the south-^south-west, and an isolated exposure of gneiss over Cambrian 
in the west reveals a tangential movement of the great gneiss flakes to the 
extent of at least 10 miles. It is obvious that on these immense surfaces 
of movement a really extraordinary amount of gliding, grinding, scoring, 
rolling out and manifold alteration of the rocks must have occurred. 

Above the edges of a system of smaller flakes, which assumes the 
form of a unilaterally compressed trough, other much larger flakes have 
been driven in an alfnost horizontal direction for a distance of at least 
10 miles. The movement came constantly from the east-south-east. Such 
then is the appearance presented by the base of the outer border of a lofty 
mountain chain, when completely levelled down by denudation. The gneiss 
of the Hebrides lies in front of it, like Bohemia before the eastern Alps. 
Ail these great movements however are older than the lower part of the 
Devonian red sandstone, and fragments of the ancient mountain chain are 
inchided in the breccias and conglomerates of this formation. 

This zone of overthrusting we will designate for brevity ' the Eriboll 
zone of averthruBt' Judging from its lie and position it corresponds to the 
great dislocation which in the extreme north separates the table-land of 
Norway from the western zone of gneiss. 

To the south-east of the Eriboll zone the Scotch highlands extend as 
fur as a line of fracture which runs from near the mouth of the Clyde 
across the south part of loch Lomond, and then across the whole country, 
keeping a north-east direction till it reaches the east coast at Stonehaven* 



80 THE OUTLINES OP THE ATLANTIC OCEAN [PARrni 

This fracture sharply bounds the highlands on the south-east, and they are 
divided in the middle by a parallel fracture which runs from loch lannhe 
right across the country, and is continued parallel to the north coast of 
the Moray firth. The first line marks the south border of the Grampians 
the second is the line of the Great Glen. The latter cuts so deeply into the 
country that it has rendered possible the construction of the Caledonian 
canal. 

The two moieties of the Scotch highlands are formed principally of 
Silurian bedi^, thrown into folds which strike to the north-east, nearly 
parallel to the zone of EriboU. These are the inner folds of the chain. 
The great fractures are therefore longitudinal fractures which run with 
the strike. On the eroded surface of these ancient mountains, on its slopes 
and in its hollows, lies the Old Bed sandstone, preserved from destruction to 
the present day, in some places as originally superposed, in others owing 
to subsidence. 

We see it, as it extends from the bottom of the Great Glen to the north- 
east, bordering both shores of the Moray firth, then broadening out over 
Caithness, and passing out to sea from this northernmost extremity of 
Scotland to form the Orkney islands. It is still visible on the south-east 
side of the Shetland group, and thence is continued, as Geikie believeSy 
along the Sogne fjord or the Dais fjord. 

South of the Great Glen granitic intrusions increase in number. In 
this region we find that the ridges which trend towards Ireland correspond 
in direction with the strike of the ancient formations which, as matter of 
fact, do indeed find their continuation in the north of that island. Finally, 
as we reach the Clyde, we encounter the faulted boundary of the highlands. 

All the coimtry south of this fracture, about as far as Girvan on the 
firth of Clyde and Dunbar on the firth of Forth, is known as the Central 
lowlands. This is the richest part of Scotland; here lie Edinburgh, 
Glasgow, and the coalfields. It is a trough; the southern edge is split 
up by numerous fractures, and consequently does not contribute so con^* 
tinuous a feature to the landscape as the northern edge. The trough is 
filled with subsided masses of Devonian red sandstone', and the various 
stages of the Carboniferous, which elsewhere have been swept away 
over vast areas, leaving exposed the Silurian horsts on the north and on 
the south. The constriction of Scotland between the firth of Clyde and 
the firth of Forth corresponds to the trough, while the horsts project to 
north-east and south-west into the sea. The subsidence would find yet 
clearer expression in the configuration of the land were it not that 
eruptive rocks, in particular porphyries, accompany the down-thrown 
sediments, and these rise in hills above their surroundings owing to the 
greater resistance they offer to denudation. 

We will now pass to a consideration of the south edge of the trough. 



CH.n] THE SCOTCH TROUGH 81 

On the sonih a second Siluriaii horst is seen, which recalls in many 
respects the highlands. The Lammermuir hills and the Moorfoot hills 
in the north-east belong to its outer part. Fractures are numerous, but 
not as it would appear continuous, and sinking movements have taken 
place here at various times. A few are certainly older, others just as 
certainly younger, than the Carboniferous period 

The principal fracture in the central region is overiaid and covered up 
towards the north-east by the Coal-measures of Midlothian. Along this 
fracture the whole thickness of the Old Bed sandstone amounting to at least 
15,000 feet is let down in the middle of the country. In this region the 
dislocation is certainly older than the Carboniferous of Midlothian, which 
rests, south-east of the fracture, directly upon Silurian; thus before the 
deposition of the Carboniferous limestone the Old Bed sandstone was not 
only faulted down, but its whole thickness was removed by erosion. Never- 
theless, subsequent movements have also affected the Carboniferous. At 
the same time, the fault in this region has so little effect on the sur&ce of 
the country that it is crossed by several river valleys ; and thus the Nith, 
a little river which rises in the southern Silurian horst, flows across the 
great fracture into the sunken region, makes a bend near New Cumnock 
in the Carboniferous, returns across the fracture into the Silurian and 
finally flows into the Solway, having thus completely traversed the 
southern horst ^ 

Thus geology points to subsidences in the midst of the mainland which 
may attain in a single fault the mean depth of the Atlantic Ocean, and 
yet this fault manifests itself so little in the landscape that a mere 
thread of water such as the Nith pursues its winding course imdisturbed 
across it. 

In the north-east, on Dornoch firth and on the south side of Moray 
firth, patches of Mesozoic and in particular of Jurassic rocks occur close 
to the sea, separated by faults from the ancient highlands behind them. 
Judd has described these rocks in detail ^. They are downthrown fragments ; 
one of them forms the Jurassic coalfield of Brora on the coast of Sutherland. 
The lines of subsidence correspond here to the course of the coast and are 
almost parallel to the Great Glen. Where the line of the Great Glen 
itself is continued along the north shore of Moray firth to the sea, some 
downthrown patches of Mesozoic lie against it. Other patches lie on 
the south-east side of the same firth. In the highlands Mesozoic deposits 
have famished abundant material to the ancient glacial d^ris, and on 
the west side, e.g. on Skye, they remain preserved from denudation 
beneath the basaltic sheets in patches of considerable size. Judd rightly 

^ Peach, Mem. Geol. Snrv. Scotland ; Ezplanaidon of Sheet 15, 8vo, 1871, pp. 7 and 37. 
' J, W. Judd, The Secondary Rocks of Scotland ; Quart. Jonm. Geol. Soc, 1873, XXIX, 
p. 118 et seq., pi. vii (cf. I, p. 206). 

■01x1911 Q 



82 THE OUTLINES OP THE ATLANTIC OCEAN [pakt in 

concludes that they must once have covered the greater part of the 
highlands and that great subsidences must have occurred even after the 
Mesozoic period. 

These downthrown patches of Mesozoic beds increase the resemblance 
of the highlands to the horsts of the Rhine, but we must observe tiiafe 
the subsidence faults in Scotland are true longitudinal faults in the 
dii*ection of the ancient folding, while in the Vosges and the Black 
Forest the folds are cut through obliquely by the faults of the Rhine 
valley K 

Shetland and the Orkneys, the Scotch highlands together with the 
trough of the lowlandis, and the southern horst, must be regarded as the 
continuation of the folded mountains of Norway. The sea which separates 
Scotland from Norway lies on a downthrown segment of these mountains, 
as is shown by the mighty fractures of the Scotch coast These pre- 
Devonian mountains, which proceed from Norway and form the whole 
of Scotland, together with their overthrust outer border along the zone 
of Eriboll, we call the Caledonian mountains. 

In Scania many movements may be shown to have taken place after 
the Cretaceous period. In Scotland we observe subsidences, some of 
which are very ancient, and others which are post-Cretaceous, probably 
even post-Tertiary, in age; Judd indeed believes that the separation of 
Scandinavia from Scotland took place in comparatively recent times, 
perhaps even after the appearance of man. A curious fact may be cited 
in support of this view. Peach and Home have found that the Shetland 
islands are everywhere polished by ice, and that the glacial striae cross the 
eastern side of the islands from the north-east, swing round to the north- 
west on reaching the central axis, and proceed in this direction across the 
western side of the islands to the Ocean. Thus a continuous sheet of ice 
coming from Scandinavia must have moved transversely across these 
islands \ 

The Caledonian mountains are continued through a great part of Ireland 
and Wales. 

Ireland includes mountains of very various structure. In the north 
rise the ridges of Donegal, readily recognized as the prolongation of the 
mountains of Scotland , the strike is the same in both. They reappear in 

> In this way Lepsins explains the Bubsidences of Saveme and in the Endchgaa; 
R. Lepsius, Die oberrheinische Tiefebene and ihre Randgebirge ; Forschungen sur 
deutschen Landea- und Volkskunde, heiauBg. v. R. Lekmann, I, 2. Heft, 1885, p. 70. 
The fanlt-lines strike out along Dornoch and Moray firth in the same manner as those 
which proceed from the west border of the Bohemian mass towards Bavaria. 

' B. N. Peach and J. Home, The Glaciation of the Shetland Isles ; Quart Jonm. 
Geol. Soc, 1879, XXXV, pp. 778-811, and geological map. The anthoxs suppose that 
the sea between Scandinavia and the Shetlands was completely displaced by the 
ice mass. 



CH. u] THE IRISH COAST 83 

\ Mayo aad on the north side of GhJway bay. In like manner the older 
fonnatioDB crop out in the aonth-eaat of the island, again with the same 
strike to the sonth'south-west ; they extend through Carlow, Wicklow, and 
Wexford, and their folding is prior to the deposition of the Old Red 
sandstone. These fragments of the pre-Devonian Caledonian mountains 
are united together by a broad platform of flat-lying Carboniferous lime- 
stcme, from the surface of which the Coal-measures have in large part been 
rHDOved l^ erosion ; it is so flat that from Dublin in the east to Qal way 
in the west there is scarcely an elevation which attains the height of 
260 feet 

In the south of the island another range of mountains appears. This is 
younger; the Old Bed sandstone and the Carboniferous limestone are here 
thrown into long folds, which strike from west to east, but are deflected in 
the south-west comer of the island to the west-south-west. The various 
antidinals run out into the Atlantic in projecting spurs, while the synclinals 
form deep bays. These submerged sjnidinals follow one another in close 
saeoeesion in the bays of Dingle, Eenmare, Bantry, Dunmanus, and Crook, 
so that the outline of the south-west coast of Ireland is entirely determined 
by folding of the crust 

On the peninsula between Dingle bay and Eenmare bay an anticlinal of 
the Old Bed sandstone rises in Macgillicuddy Beeks to form the highest 
peak in Ireland (3,414 feet). Near the lake of Eillamey this anticlinal is 
completely overfolded to the north, so that the flat-lying Carboniferous 
limestone of central Ireland is found here together with pinched-in Culm 
beds, dipping to the south, beneath the Old Bed sandstone. 

These folds and overf olds in the south of Ireland also strike out to sea 
on the east coast of Cork and in Waterf ord. They are only a part of 
a great arcuate range, which is continued from here towards the east into 
Wales and England, and thence still further into Belgium \ 

This great arc, striking from east to west, and opposed in so marked 
a manner to the older Caledonian folds striking to the south-south-west, 
we will designate ihe Armorican arc : we here meet with its northern border 
for the first time: the contrast between it and the Caledonian folds is 
eepedaQy weU seen in the south-east of Ireland, as in the counties of 
Kilkenny and Carlow. 

Let us now return to the other side of St. Oeorge's channel. MoQpre has 
shown how in Wigtownshire^ on the south-west coast of Scotland, the 
rocks of the Southern Uplands, thrown into steep folds, proceed towards 
the sea between Corswell lighthouse and the Mull of Galloway. In the 
north of this coastal region the greater part of the folds are overturned to 



' Hull, Physical G^logy and Geography of Ireland, p. 188. A section through 
Di^le, Kenmare, and Bantry bay, showing a great fracture in the first, and synclinals 
11 tbe other two bays, is given by Einahan, QeoL Mag., 1879, 2nd ser., VI, p. 851. 

G a 



84 THE OUTLINES OF THE ATLANTIC OCEAN [part hi 

the north-west ; in the south, on the other hand, to the south-east ; so that 
an imbricated fan structure is produced ^. 

In like manner the gneiss ridges of Anglesey strike across the island 
to the south-west, and are continued on the south coast of Carnarvon bay 
into the peninsula of Ueyn, which faithfully follows the Caledonian 
direction. These parts of North Wales belong, however, as shown by their 
strike, to a part of the Caledonian region lying further to the east than 
the ranges of Carlow and Wexford mentioned above ; these, it is true, 
are followed in the extreme south-east of Ireland near Camsore point, 
south of Wexford bay, by another small exposure of ancient rocks. 

The folds of North Wales, like those of Carlow and Wexford, are 
older than the Old Bed sandstone. From the north side of Anglesey 
the Carboniferous limestone extends in a great arc along the border 
of the folded region, so that throughout the east of Wales it rests on the 
various members of the Silurian formation; it is accompanied by a dis- 
continuous band of Old Bed sandstone which further south broadens out 
over a wide area in Hereford and Brecknock^. 

We have now reached South Wales, that region in which we may see, 
thanks to profound erosion, the encoimter of the pre-Devonian Caledonian 
ranges, striking south-south-west to south-west, with the more recent post- 
Carboniferous Armorican folds, striking east to east-south-east. Here 
I follow principally the description given as early as 1846 by De la Bdche ; 
this I cannot mention without an expression of deep gratitude to the 
author, now long since dead, since it exercised many years ago a decisive 
influence on my own views as to the structure of great mountain ranges. 
Although published forty years ago, this description is inspired by those 
conceptions of the formation of mountain ranges by lateral pressure, and 
of the true influence of granite masses, which are now winning their way 
step by step to general acceptance ^. 

The boundary line between the Caledonian and Armorican regions may 
be recognized with comparative ease on any geological map, for the north 
border of the Armorican range coincides with the south border of the coal- 
field. Leaving Waterford in Ireland, the boundary line enters St. Bride's 
bay, Pembrokeshire, with a strike bent a very little from east to east- 
south-east, passes across the upper part of Caermarthen bay, then across 
Swansea bay to the broad estuary of the Severn which it reaches in the 
neighbourhood of Cardiff*, it crosses the Severn and is continued on the 

^ J. C. Moore, On the Silurian Bocks of Wigtownshire ; Quart. Joum. GeoL Soc., 
1856, XII, pp. 359^866. 

^ A« C. Bamsay, The Geology of North Wales ; Mem. Geol. Surv. Great Britain, 
vol. ni, 186&, p. 13. 

' Sir Henry T. de la B^che, On the Formation of the Bocks of South Wales and South- 
west England ; Mem. Geol. Surv. Great Britain, 1846, vol I, pp. 1-296, in particular, 
pp. 221-239 ; cf. Ihitstehung der Alpen, p. 16. 



CH. n] THE COASTS OF WALES 86 

north border of the Hendip hills. Thus it is only the southern part 
of the three promontories on the north coast of the Bristol channel which 
belongs to the Armorican folds. North of the boundary, the rich coal- 
fields of St. Bride's bay extend in the form of a narrow compressed zone 
to Gaermarthen bay; this they cross, and attain on its further side 
a continually increasing breadth until they spread out in Glamorgan 
and Monmouth as the great coalfield of South Wales. This expansion 
correeponds to the retreating curve of the northern mountains — the 
continuation of the Caledonian hill-ranges of North Wales ; these we must 
DOW consider in greater detail. 

In North Wales we have already become acquainted with the typical 
Caledonian strike in Anglesey and the peninsula of Ueyn, and we have also 
seen that the Caledonian folding is older than the Old Bed sandstone. 
The Caledonian folds are continued through Merioneth and are sometimes 
directed from north to south, but in Cardigan they make a bend to the 
south-west which becomes more and more marked towards the south, and 
beyond the town of Cardigan towards Fishguard they swing completely 
round into the Aimorican direction, or from east to west ; the beds are then 
very much disturbed both in a horizontal and vertical direction. Here, 
' in the northern part of Pembrokeshire/ says De la BSche, ' we may have 
the complicated forms resulting from a twist of the rocks in a new 
direction over an older one \' 

St. David's head, which bounds St. Bride's bay on the north, consists, 
according to A Geikie, of an anticlinal of Cambrian rocks, which strikes to 
tiie south-west, that is obliquely across the peninsula, and is overfolded to 
the south-east K The predominance of a trend so entirely different from 
that of the Armorican folds is very remarkable. Towards the interior of 
the country, it is true, the overturned south-east side assumes a vertical 
position; but towards the head of St. Bride's bay, next the pinched-in 
band of Coal-measures, great fractures and disturbances occur in the 
Cambrian rocks ', and it is not till we proceed further to the east, where 
the coalfields of Glamorganshire broaden out, that we first see with some 
distinctness a deflexion from the Caledonian to the Armorican direction^ 
Here, according to De la Bdche, the great coalfield forms three anticlinals 
which strike to the south-west, but towards the south, these are succeeded 
by others which strike from east to west. This might well be a case of 
syntaxis, such as we have already studied in the anticlinals of the foot- 

' Op. cit., p. 223. 

' A. Gteikie, On the suppofied Pre-Gambrian Rocks of St. David's ; Quart. Joum. Geol. 
See., 1883, XXXISi, pp. 261-333, map, p. 268 ; — granite crops out from the overturned 
limb. On the other hand Hicks, On the Pre-Cambrian Rocks of Pembrokeshire ; op. 
dt, 1884, XL, pp. 507-560, map, where the presence of Archaean rocks is maintained. 

* J. £• Marr and T. Roberts, The lower Palaeozoic Rocks of the Neighbourhood 
of Haverfordwest ; Quart. Joum. GeoL Soc., 1885, XLI, pp. 476-491, pi. xv. 



86 THE OUTLINES OF THE ATLANTIC OCEAN [pabt m 

hills of the Himilaya and of the Hindu Enah on the JehlanL But there 
are other tectonic features, particularly in Ireland, which show that we 
have here, not a case of syntaxis as on the Jehlam, but the encounter 
of a younger arc with an older folded system, resembling rather the 
encounter of the Carpathians and the Sudetes in Silesia and north Moravia. 

We will not enter in detail into the structure of Hereford and Glouoester, 
but turn our attention to the south. 

7. The Armorican mowrUains. We have seen that in the south of 
Ireland the Devonian and Carboniferous beds are thrown into great folds, 
which in the south-west of Cork strike to the west-south-west, but 
elsewhere, in Cork, Kerry, and Waterford, from east to west, these folds are 
sometimes overfolded to the north, and they crop out again towards the 
east in South Wales. We there traced their northern boundary from 
St Bride's bay, along the south border of the great coalfields, across 
Caermarthen bay and Swansea bay to the neighbourhood of Cardiff on 
the lower Severn. This line is distinguished near Tenby in Caermarth^i 
bay, and in Swansea bay, by overthrust inversion of the Coal-measures ; 
all the country visible to the south of it, namely the three peninsulas on 
the north side of the Bristol channel, consists, like the folds of Ireland, 
of antidinals and synclinals of Old Bed sandstone and Carboniferous lime- 
stone, in so far as they are not concealed by patches of the transgressive 
covering of younger sediments which begins here. This transgressive 
series commences with the Permian and includes more to the east the 
whole Hesozoic series. 

The direction of the northern boundary of the Armorican system and of 
its folds themselves is here east and a little south, and the continuation of the 
bands of Carboniferous limestone across the Bristol channel is marked by 
rocky islands, such as the Steep Holme, for example. This islet was rightly 
recognized by Buckland and Conybeare, in the year 1824, as the fragment 
of a fold overturned towards the north. These keen-sighted observers also 
perceived that the island of Flat Holm, lying a little to the north of the 
Steep Holme and likewise formed of Carboniferous limestone, is a flat 
anticlinal striking north-east towards the peninsula of Brean Down ; we 
can therefore no longer assign it to the Armorican folds K 

^ W. Buckland and W. D. Conybeare, ObaervationB on the South-weit Goal Digtoict 
of England, Tnuu. Geol. Soc, 1824, 2nd ser., II, pp. 214-282 ; figure of the inverted 
stratification on Steep Holme, p. 283. The peculiar nuuiner in which the mountain 
chains meet, the rectilinear, south-westerly strike in north Wales, and the bend to the 
south-west which the Silurian zone presents near St. Bride*s bay may be seen on 
Ramsay*s general map, Geology of North Wales, Mem. Qeol. Sunr., 1866, yoL III. 
A slight deviation in north Wales from a south-west or south-south-westerly direction 
to one somewhat more southerly is shown in the diagrammatic sketch of the directions 
of strike given by Larivi^re, Notes d'un voyage auz ardoisi^res du Pays de Galles, Ann. 
des Mines, 1884, 8« s^r., VI, pi. ziii, fig. 1. 



CH. n] THE BRISTOL CHANNEL 87 

Thus the Steep Hohne, situated north of Bridgwater bay, leads ns to the 
long antLcIiual of the Mendipe, which extending from the sea-coast towards 
Frome forms the most northerly of the Armorican antidinals. It consists 
of Carhonif erous limestone through which the Old Bed sandstone makes its 
appearance in four exposures. In its eastern part the arching passes into 
overfolding to the north; the Coal-measures lying in front of it on the 
north are crumpled up in zigzags, and dip towards the south beneath the 
anticlinal of Carboniferous limestone. The lie of the beds bears so strong 
« reeeEmblance to that of the Belgian Coal-measures, also overf olded from 
ihe south, that, as early as 1824, Buckland and Conybeare, availing 
themselvee of the descriptions given by Omalius d'Halloy, compared' the 
eastern part of the Mendips with the neighbourhood of Namur and Liittich \ 

Near Frome the Carboniferous disappears beneath the Cretaceous, but 
the anticlinal of the Mendips is continued, as we shall see, in the form 
of a much more recent anticlinal towards the east. South of the Mendips, 
at the head of Bridgwater bay, lies alluvial land with patches of Mesozoic, 
bat soon there rises in the south-west the extensive Devonian region which 
forms west Somerset, Devonshire and Cornwall. 

This great peninsula belongs entirely to the Armorican folds, and the 
strike of the Devonian and Carboniferous beds, of which it is chiefly formed, 
corresponds precisely, especially in the northern region, to that of the 
Mez&dips. It presents on the north a zone of Devonian rock, which strikes 
from Bridgwater and the Quantock hills towards Lundy island, the 
northern half of this island being formed by an intrusion of more recent 
granite ; then follows a broad central zone, in which the Culm beds attain 
a <x>nfiiderable development, and last a southern Devonian zone, which 
includes the greater part of the south coast of the peninsula and the 
southern half of its Atlantic coast. This southern Devonian zone is marked 
by a series of great post-Carboniferous granite bossed. The most easterly 
and extensive of these forms Dartmoor forest, south-west of Exeter; a 
second lies west of Liskeard, a third north of St. Austell, a fourth west of 
Falmouth ; the larger mass which follows forms the extreme end of the 
peninsula as far as Land's End, and the Scilly isles are the summits 
of another granite mass covered by the sea K 

' Torn, cit., p. 220 ; for the Mendips in particular, G. Moore, On Abnormal Gonditiona 
of Secondaiy Deposits when connected with the Somersetshire and South Wales Coal 
Basin, Quart. Joam. GeoL Soc, 1867, XXIII, pp. 449-568, in particular p. 451 et seq. ; 
H. B. Woodward, Geology of the East Somerset and the Bristol Coalfields, Mem. Geol. 
But. England and Wales, 1876, pp. 22, 190 et seq. Attempts to explain the inversion 
of the Coal-measures by the method adopted in this work have only led to fieur greater 
complications ; cf. H. B. Woodward, Remarks upon Inversions of Carboniferous Strata 
in Somersetahire, GeoL Mag., 1871, YIII, pp. 149-154. 

* A. Sedgwick and R I. Murchison, On the Physical Structure of Devonshire ; Trans. 
<jeol. Soc 1840, 2nd ser., Y, pp. 638-704, pL 1-lviii, map. The Devonian of north 



88 THE OUTLINES OF THE ATLANTIC OCEAN [pabt m 

The granite masses cause considerable local disturbances in the strike, 
which otherwise faithfully follows the curve of the Armorican folds. From 
the west of liskeard down to the Scilly isles the series of granite masses 
is directed more to the south-west than we should have expected from 
the strike of the folds in the Bristol channel, and it has therefore been 
supposed that they are independent of the strike. But the slight deviation 
to the south-west corresponds to the similar deviation in Cork, e.g. in 
Bantry bay, and this shows that the vertex of the Armorican arc lies between 
Ireland and Wales or in Pembrokeshire. Further we shall see soon that 
the two granite masses of the Harz, which are precisely analogous to those 
of Cornwall, are by no means disposed according to the strike. De la BSche 
conjectured the existence of a subterranean connexion between the granite 
masses, and with admirable sagacity recognized, as early as 1846, that 
it was not by their eruption the mountains had been upheaved, but as 
a consequence of a much more general force which at the same time 
manifested itself in folding. 

South of the series of granite masses, along the south edge of the 
southern Devonian zone, Archaean rocks crop out on the south coast 
in two restricted areas at the Lizard in the west and Prawle Point 
(south of Dartmouth) in the east. These rocks form the foundation o£ 
the Devonian mountains, though in the south these also include Silurian 
deposits. In the first locality, or the Lizard, hornblende schists and 
serpentine are visible; in the second at Prawle Point, chloritic schists and 
mica schists. At the Lizard the strike is south-west, corresponding to that 
of the south-west of Ireland and of the series of granite masses ; at 
Prawle Point it is directed east and west Between these two places ree& 
of rock rise -from the sea, one of which bears Eddystone lighthouse: 
these consist of gneiss. 

Let us now attempt to obtain a general view of the mountain fragment 
of south-west England. The most ancient foundation is seen in the south 
as the gneiss of the Eddystone. This is followed by fragments of an 
ancient mantle of schists, the hornblende schists of Lizard Point in the 
west, and the chloritic and mica schists of Prawle Point in the east. Thus 
we reach the great southern zone of Devonian, with subordinate bands- 
of Silurian, which encloses the long series of granite masses extending 
from the Scilly isles to Dartmoor; then further to the north follows 

Devon, which forms the south coast of the Bristol channel, is formed of beds regularly 
inclined to the sonth, which dip beneath the Culm of the synclinal, bat immediately in 
front of the east boundary of the horst they experience, evidently owing to a local: 
advance of the mountains, a sigmoid flexure of the strike towards the north, and 
in front of this east boundary stand, as a fairly independent fragment, the Quantock hillsy 
in which the elbow is continued. But the return to the normal direction of the 
Mendips takes place deep below the ground and is invisible. These relations are very 
clearly shown on the map by Etheridge, Quart. Joum. Geol. Soc, 1867, XXIII, p. 580. 



CH. n] THE COAST OP BRITTANY 89 

the broad Culm zone, then the north Devonian zone as far as the Bristol 
channel. At the same time all these mighty deposits are thrown into 
narrow parallel folds, which are frequently overturned to the north, 
esp^sially in Somerset. Then still further to the north there follows the 
anticlinal of the Mendips (beneath which the inverted Coal-measures dip 
to the south), and the western prolongation of the Mendips, which extends 
as far as St. Bride's bay. 

Such is the structure of a great folded complex, developed under 
a pressure acting from the south. With perfect justice Bonney remarks 
that this great series of folds scarcely yields in importance to the existing 
chain of the Alps, and he compares the gneiss of the Eddystone with the 
gneiss cores of the Alps \ What we see here is the ruins of a mighty 
mass of lofty mountains ; the Lizard and Prawle Point, the last remains 
of the schist zone, project like cape Matifou and the peninsula of Bonzar^ 
on the two sides of the bay of Algiers (I, p. 228). 

Bonney believes that the foundation of this great mountain region is 
continued in the ancient rocks of the neighbouring parts of northern 
France, and to these we will now turn our attention. 

The north-west of France is formed of Archaean and Palaeozoic rocks ; 
they form the soil of Brittany and extend from the Cotentin into the 
Vend^ Towards the east and south their continuation is concealed by 
Jurassic beds. The eastern boundary of their outcrop runs from the east 
side of the Cotentin to Alenfon, and thence southwards through Angers 
to Fartenay and Saint-Maixent, north-east of Niort. In this region the 
boundary turns in a right angle to the west, and reaches the sea north 
of La BocheUe, near Les Sables d'Olonne. 

The strike of these older rocks is directed to the west-north-west with 
deviations to west and north-west, and in the western parts of Brittany 
the westerly direction appears to predominate. A glance at the excellent 
but somewhat antiquated map of £lie de Beaumont and Dufr^noy shows 
OS that the. peninsula of the Cotentin is a fragment of a great horst, like 
the Morvan or the Thiiringerwald, and that the strike runs transversely 
across it from east to west, while the lie of the great Armoiican peninsula 
eorresponds with the strike. 

Much has been written on this part of the country: Dalimier has 
described the Cotentin; for Normandy I may mention the works of 
Barrois, Hubert and L^besconte. It is above all Barrois who has devoted 
particular attention to the tectonic problems, and, supplementing the earlier 
investigations of Boblaye, has elucidated the structure of the country K 

^ T. G. Bonney, On the Oeology of the South Devon Coast from Torcross to Hope 
Cove ; Quart. Joum. Geol. Soc, 1884, XL, pp. 1-25, in particular p. 24. 

* P. Dalimier, Stratigraphie des terrains primaires dans la presqu^e du Cotentin, 8^, 
^KOB, 1871 ; C. Barrois, Observations sur la constitution g^ologique de la Bretagne, Ann. 



90 THE OUTLEJES OF THE ATLANTIC OCEAN [pabt in 

The whole region, extending through three degrees of latitude, is 
thrown into nearly parallel folds. Here also the most important folding 
occurred within the Carboniferous period. 

There are two great antidinals and a number of subordinate ones. The 
northern of the two great antidinals forms the peninsula to the north of 
Brest, with the Montagues d'Arrfo, and extends as far as the Jurassic near 
Alenfon, its northern flank nearly oorndding with the north coast of 
Brittany; it consists chiefly of Archaean rocks, which are frequentiy 
traversed by granite. The southern great antidinal is formed of Cambrian 
beds, and is penetrated in many places by granite ; it forms the Pointe du 
Raz, the most southerly of the three spurs which Brittany sends out 
towards the Ocean, comprises the Morbihan, and runs past Vannes and 
Nantes. Between these two great antidinals and to the north and south 
of them, numerous folds, crowded together, run side by side almost in the 
same direction ; in these also, e. g. near Bostrenen, north of the Morbihan, 
post-Carboniferous granites have been intruded. 

It is a significant fact that the first of the northern syndinaLs, the 
Bassin de Mortain, which crosses the Cotentin peninsula north of Alenfon, 
passing through Domfront and Mortain, does not terminate on the coast 
in the Bale de Cancale/but according to Barrois it reappears more to the 
west on the north coast of Brittany at cape Frehel, and then, keeping the 
same direction, runs towards Paimpol and traverses the northern extremity 
of Brittany. 

On the west coast, the northern antidinal is continued into the island 
of Ouessant, and ihe southern passes from Quimper across Pointe du Bas 
into the little lie de Sein ; between these two great antidinals there lies 
a broad synclinal, which indudes the various members of the Palaeozoic 
series from the Silurian to the Carboniferoua It embraces many sub- 
ordinate folds, and extends from Laval on the Mayenne westwards to the 
sea. To this prindpal synclinal belongs the irr^ular-shaped spur which, 
south of Brest, projects towards the Ocean between its two companions on 
the north and south. 

The whole country is thus folded in the same direction as the south 
of Irdand and England, and speaking broadly the folding was acoom- 

6oc. g^oL Noid, 1883-1884, XI, pp. 87-91 and 278-285 ; by the same, La staracture 
etratigraphique des montagnes du Menez, op. cit., 1885, XIII, pp. 65-71 ; by the same, 
Aper^u de la Btructure g^Iogique dn Finistfere, Bull. Soc. g6ol. de Fr., 1886, 8^ 8^r^ 
XrV, pp. 655-665, and Aper9u de la constitution g^ologique de la rade de Brest, torn. 
dt., pp. 678-706 et passim ; £. Hubert, Phyllades de 8aint-L6 et conglom^rats pouipr^ 
dans le Nord-Ouest de la France, torn, cit., pp. 713-774, and Observations snr les groupes 
s^dimentaires les plus anciens du Nord-Ouest de la France, Compt. Rend., 1886, CIII, 
pp. 230-285, 303-308, and 367-371 ; P. L^besconte, Constitution g^n^rale du massif 
breton compar^e k celle du Finist^re, BnlL Soc. g^l. de Fr., 1886, 3* s^r., XIY, pp. 776- 
819 et passim. 



CH. n] THE BELGIUM POREFOLDmG 91 

pliahed in both regions at the same time. In breadth this great folded 
region, now broken up, extends from the Mendips into the Vend^^ 

In* the neighbonrhood of Exeter, where the peninsula becomes narrower, 
the Palaeozoic formations dip beneath a Mesozoic covering, and more to the 
ix>rth, near Frome, the anticlinal of the Mendips disappears in the same 
manner beneath the more recent sediments. This mantle of younger rocks 
ooneeals the foundation of the whole south-east of England But we have 
abeady mentioned that the similarity in the lie of the beds on the north 
border of the Mendips and in the Belgian coalfield is so great that it 
was observed many years ago by Buckland and Conybeare. In 1865, 
Cbdwin Austen ventured to express the opinion that an actual connexion 
exists below the ground between the overtiirust Coal-measures of the 
Mendips and those of Boulogne, and that it might even be possible by 
means- of boring to prove the existence of these Coal-measures in the 
n^hbourhood of London itself. He based this view not only on the 
correspondence in the lie of the beds and the strike of the folds in these 
two regions, so far distant from each other, but also on the presence of 
more recent disturbances in the superposed Mesozoic beds, which follow the 
same strike \ 

The borings which have been made have revealed various Devonian 
beds at a great depth, and it has been placed beyond doubt that beneath 
LondoQ the Great Oolite rests directly, without any intervening lias or 
Trias, upon Palaeozoic beds, as is also the case near Calais, where the 
continuation of the Belgian f oref olding reaches the sea ^. This f oief olding 
has been discussed in an earlier passage (I, p. 142 et seq.). Along a line 
ronning from Boulogne to Aix-la-Chapelle older beds are driven forward 
over younger. Bertrand compares this zone of forefolding with certain 
ports of the Alps of Glarus^. There is indeed the greatest resemblance 
to ihe overfolded outer borders of the Alps, the Carpathians or the 
Himalaya. That this zone of forefolding finds so little expression in the 
eoofigiiration of the country, and that its presence in the north of France 
was only discovered by means of borings, has no bearing on the immediate 
question. All that has been established in this region by mining or by 
laborious comparative studies finds its sole analogy in the outer borders 
of the greatest mountain chains now existing, or in that ancient border, 
exposed by profound abrasion, as the f oref olded zone of Eriboll. 

A difference presents itself, however, in one striking particular. The 

^ K, Godwin Austen, On the possible extension of the Coal-measures beneath the 
8oiitk-eai(tem part of England ; Quart Joum. Geol. Soc., 1856, XII, pp. 88-78, map. 

* Jndd, On the Nature and Relations of the Jurassic Deposits which underlie London ; 
«p. cit, 1884, XL, p. 754. 

' M. Bertrandy Rapport de structure des Alpes de Glaris et du bassin houiller du Nord; 
BqIL Soc g^oL de Fr., 1884, 8« s^r., XII, pp. 818-880, pi. xi. 



92 THE OUTLINES OF THE ATLANTIC OCEAN [part in 

outer borders of the great moantain chains are always more or less convex 
in the direction of the tangential movement. A concavity or even 
a re-entrant angle within a region of forefolding has always been regarded 
as an indication of syntaxis, i. e. of the encounter of two differently directed 
components of the folding force. 

The Belgian zone of forefolding is concave. The descriptions given by 
Oosselet and Delwaque's map show this plainly \ From Calais to Douai 
the overthrust zone strikes E. IS"* S. ; between Douai and Valenciennes it 
is abruptly bent and extremely disturbed ; from Valenciennes it turns first 
east-north-east and then near Aix-la-Chapelle to the north-east. The 
western part is thus folded from south-south-west to north-north-east; 
the eastern, on the other hand, from south-south-east to north-north-west 
and from south-east to north-west, while between Douai and Valenciennes 
lies the syntaxis in a gentle curve. 

Thus the outer borders of two vwuTUain systems vmite in the Belgian 
zone of forefolding. The eastern part is the outer border of the Ardennes, 
and the western is the continuation of the Mendips or the outer border 
of the Armorican arc. The Armorican arc itself is at this point almost 
concealed by the overlying Mesozoic sediments. The presence of Palaeozoic 
beds has, it is true, been established by means of boring in many places 
west of the coal-bearing zone of Calais-Douai, but it is only near Marquise, 
north of Boulogne, that the Devonian rises in a small outcrop from 
beneath the Jurassic mantle. From this place the strike passes across 
to England. 

The folds of the south of Ireland, those on the north side of the Bristol 
channel, the anticlinal of the Mendips, the folded region of Somerset, 
Devon, and Cornwall, the ancient folded mass of the Cotentin, Brittany, 
and the Vend^ ; then, to the east, the Devonian island of Boulogne and 
the western half of the forefolded zone of Belgium in its course from 
Calais to Douai ; all these we now see unite together to form the great 
Armorican arc. A broad gap however remains between Exeter and 
Boulogne. 

The formation of this arc took place towards the close of the Carboni- 
ferous period ; the great mountains were then worn down by denudation, 
and covered unconformably by thick deposits of more recent sedimente. 
Then piece by piece the mountains collapsed, and one of the most extensive 
subsidences is that between Exeter and Boulogne. It appears, however, 
that within the limits of this Bulmdejice fresh foldirigs subsequenMy occurred 
which followed the ancient Armorican direction. 

In the region indicated above, between Exeter and Boulogne, lies the 

* J. Gosselet, Esqmsse g^ologiqne dn Nord de la France, fasc. Terraiiui piimairea,, 
8<>, Lille, 1880, atlas; G. Dewalqne, Carte g^logiqne de la Belgique, folio, Bnusels, 
1879. 



n] BOULOGNE 93 

denuded dome of the Weald. Bounded on all sides by the escarpment 
of the Chalk, it rises over an elliptical area, which includes the greater 
part of Sussex and a part of Kent, and extends some distance beyond 
Bonlogne. 

Hopkins showed in 1841 that this dome consists of several parallel 
antidinals— three to four, at least — closely crowded together; in the west 
they strike east to west, but in the east they are deflected to the south-east, 
in the direction of their prolongation into the Boulonnais K 

In France it was found by d'Archiac, in 1846, that the watershed of 
Artois marks an important boundary line in relation to the development 
of the Cretaceous system, and that this line, which he called the * axe de 
rArtois,' runs to the north-west (W. 84'' N.) from Arras, but that in the 
Boulonnais it undergoes a deflexion to the west, and is continued beyond 
the Channel into the Weald ^ The axis of Artois, however, corresponds 
with the zone of overthrust Coal-measures, which, as we have seen, forms 
the outer border of the Armorican folds; on its site borings have been 
made which reached Palaeozoic beds, and the Devonian outcrops in the 
neighbourhood of Boulogne already stand within the Cretaceous escarp- 
ment which extends from England, and bounds the antidinals of the 
Weald. 

As early, therefore, as 1846 a number of folds were known, having 
the form of a circular arc, which run first to the west-north-west, then 
curve across the Channel and extend in a more and more east to west 
direc^tion through the south of England 

The significance of the axis of Artois for a knowledge of the English 
folds has been fully recognized by Godwin Austen, and formed one of the 
most important arguments in support of his theory that the coalfields of 
Mods and the Boulonnais must find a subterranean continuation beneath 
the neighbourhood of London and as far as the coal districts north of the 
Mendips. Indeed, Godwin Austen in his now famous treatise on this 
subject even maintained as a universal law that when any zone of the 
earth's crust is considerably folded or fractured, subsequent disturbances 

1 William Hopkms, On the geological Stmcture of the Wealden District and of the 
Bas-BonlonnaiB ; Trans. Greol. Soc, 1845, 2nd ser., VII, pp. 1>51, map. The meridional 
tnuDsvene firactnres described hj Hopldns are partly true flaws, i. e. dislocation planes 
Hke the Medina &nlt in the isle of Wight. The important earthquake of April 22, 
1884, -would thus be a flaw-shock, like so many Alpine earthquakes, and an indication of 
the penistence of forces similar to those acting in the Alps; Meldola and White, 
Nature, January 21, 1886, p. 265. 

* D'Archiac, Etudes snr la formation cr^tac^e des versants sud-ouest, nord et nord- 
ouett du Plateau Central de la France, M6m. Soc. g^oL de Fr., 1846, 2« sdr., II, p. 116 ; 
on the map pL 1, a dotted line shows the supposed connexion of the axis of Artois with 
the Weald. At that time the Coal-measures were supposed to be overthrust near Arras^; 
this coigecture was conflrmed by borings ; Degouss^e and Laurent, On the Valenciennes 
Coal-Basin, Quart. Joura. Qeol. Soc, 1856, XII, p. 252, pL t. 



94 THE OUTLINES OP THE ATLANTIC OCEAN [pabt in 

follow the previous lines, and this simply because these lines appear to be 
lines of least resistance \ 

The anticlinals of the Weald were described later in great detail by 
Topley, and their deflexion towards the Boulonnais may be regarded as 
completely established ^ 

But these anticlinals are continued towards the west across the r^on 
of the Weald ; and in addition a new and mighty anticlinal succeeds in 
the south, which again brings the deposits of the Weald into view^ but 
lies for the greater part beneath the sea. It is visible in the southern half 
of the isle of Wight; its northern flank with deeply upturned, and in part 
even inverted, beds runs through the whole breadth of the island from 
the Needles in the west to Culver cliff in the east, and the rhomb-like 
form of the island is determined by the course of the rocks which offer 
the greatest resistance to denudation ^ From here it continues to the west 
across Purbeck to Weymouth ; and south of Weymouth, on the peninsula 
of Portland, we meet with Jurassic beds which are steeply upturned. 

While the exploration of these folds was being pursued in Elngland, it 
was discovered in the north of France that a series of similar anticlinals 
striking to the north-west exists outside the axis of Artois and parallel 
to it. They have this character in common with the English folds, tibat 
the dip of the northern limb of the anticlinals is steep, while that of the 
southern limb is very gentle. The most important of these is the line 
of disturbance of the Pays de Bray, which in the midst of the basin of 
Paris, between Beauvais and Neuchatel, exposes not only the lower 
members of the Cretaceous but also the upper Jurassic. This remark- 
able disturbance was first described by £lie de Beaumont and then in 
great detail by De Lapparent^. Hubert and De Mercy have made known 
to us yet other folds, also striking to the north-west, but of less impor- 
tance, and Hubert even believed that he could prove the existence of 
a second system of disturbances running at right angles to the firsts 

^ Godwin AuBten, torn, cit., p. 62. The folds had been already described by 
P. J. Martin, On the anticlinal Line of the London and Hampshire Basin, FhiL Mag., 
1851, 4th ser., II, pp. 41-51, 126, 189, 278, 366, 471 ; further, op. cit, 1856, 4th ser., XII, 
pp. 447-452 ; and 1857, XIII, pp. 33, 109. 

■ William Topley, The (Geology of the Weald, Mem. GeoL Surv. England and Wales, 
8yo, 1875, maps ; in particular, pp. 216 et seq. 

* Edward Forbes, On the Tertiary FluTio-Marine Formation of the Isle of Wight ; 
Mem. GeoL Surv. Great Britain, 8vo, 1856, map. 

^ A. de Lapparent, Note sur le soulevement du Pays de Bray et rouTertare de 
la Tall^ de la Seine; BulL Soc. g^L de Fr., 1871, 2« s^r., XXIX, pp. 231-288, pL 1 ; 
and by the same, Traits de G^ologie, 2« ^d., 8^o, Paris, 1885, p. 1420. 

' Hubert, Note sur la craie blanche et la craie mameuse dans le bassin de Paris, BulL 
Soc. g^oL de Fr., 2« s^r.,1863, XX, pp. 605-681 ; Ondulations de la craie dans le bassiii 
de Paris, op. cit., 1871, 2« s^., XXIX, pp. 446-472, and pp. 583-595 ; farther, op. cit, 
1875, 2« s^.. Ill, pp. 512-546, map ; by the same, Ondulations de la craie dans le Nord 



CH. n] POSTHUMOUS FOLDING 96 

dually, C. Barrois, after a careful study of the numerous works of his 
piedeoeBBors and a detailed examination of the folds in England which 
trsTerse ihe Chalk to the west of the Weald, succeeded in obtaining the 
following results^: — 

The axis of Artois leaves France somewhat south of the older rocks 
of the Boulonnais, reaches England at Dungeness, is continued as one of 
tiie antielinals in the northern part of the Weald, and proceeds with an east 
to west direction through the Chalk past Eingsdere towards Ham, Frome, 
and the Bristol channel. 

A second line, designated by Hubert the axis of La Bresle, strikes to 
the north-west, reaches the sea near Tr^port, and coincides with that great 
■ anticlinal of the Weald which Hopkins calls the line of Qreenhurst ; this 

I sbrikee past Petersfield and Winchester through the Chalk, past Stockbridge 

into the valley of Warminster. 

The axis of the Pays de Bray, finally, distinguished by the great extent 
of the dislocation, must probably be regarded as part of that great fold 
which runs through the isle of Wight to Purbeck. 

Each of these folds is steeper towards the north than towards the 
soath; in some of them the folding movement has certainly continued 
far into the Tertiary period. 

These results are far reaching in their significance. Even if it should 
be shown later that some of the lines in question do not traverse the 
whole distance as continuous folds, but that contiguous antidinals running 
in tiie same direction replace each other, as in the Jura, yet this will not 
affect the fact that there exists a system of folds formed under a move- 
ment to the north-east and north, which strikes to the north-west in 
France, curves round in an arc to west-north-west and west in the region 
of the Channel, and extends with a westerly strike through the south of 
England to Weymouth and the Mendips. These lines correspond, however, 
to tiie downthrown segments of the Armorican arc and join together the 
projecting horsts. The region was folded, as we have seen, at the close 
of the Carboniferous period, was covered with younger sediments and 
subsided ; then there occurred in the same place a folding of the younger 
sediments, and this more recent folding coincides in direction with the 
6Lder folding which preceded it. 

This phenomenon we term posthumous f elding. It is very likely that 
in most othw mountain systems repeated movements in the same direction 

de la Franoe, Ann. ScL g^L, 1876, VII, Art. no. 2, 48 pp., map ; N. de Mercej, Note sur 
la ciaie dans la Noid de la France, Boll. Soc. g^ol. de Fr., 1868, 2« a^r., XX, pp. 
631-644. 

^ GL Barrois, Ondnlations de la craie dans le Sud de TAngleterre, Ann. Soc. g^l. 
Hold, 1875, m, pp. 85-111 ; and in particular, Recherches sur I9 terrain cr6tac6 
tup^rieiir de I'Angleterre et de Tlrlande, 4to, Lille, 1876, pp. 114-128. 



f 



96 THE OUTLINES OF THE ATLANTIC OCEAN [pakp in 

have occurred at very different times; but seldom do we witness so striking 
an episode as is here presented in the subsidence of a great segment of 
a mountain arc between successive periods of folding ; and in this example 
we find clearly displayed the extraordinary constancy in the direction of 
the folding force. 

The result then of all the observations we have now brought together 
is as follows : — The coast of the Atlantic Ocean, from the mouth of the 
Shannon to some distance beyond the mouth of the Loire, is formed by the 
breaking-off and subsidence of fragments of a great mountain chain, which 
was folded towards the north, and acquired the greater part of its elevation 
towards the close of the Carboniferous period ; its vestiges can now only 
be traced in isolated horsts and in posthumous foldings. The oldest rocks 
of this chain may be seen in the Yendde, Brittany, and the Cotentin, 
as well as in the gneiss of the Eddystone, and the ancient exposures of 
schist on the south coast of the peninsula of Cornwall. The next zone, 
consisting chiefly of Devonian deposits, and marked in its southern part by 
numerons granite maaaea. forma the peninsula of ComwaU and Devon 
up to the Bristol channel. The outermost zone, overthrust towards the 
exterior, is for the most part of Carboniferous age, but beneath it the 
underlying Devonian also crops out. Its northern boundary, which at the 
same time represents the outer border of the chain, runs across the south 
of Ireland to St. Bride's bay, continues close to the north coast of the 
Bristol channel and along the north side of the Mendips as far as Frome, 
then, indicated by posthumous movements, it follows close to the north 
flank of the W^den dome, the North Downs, and with a gradually 
increasing curvature to the south-east passes between Boulogne and Calais, 
and is thence continued beyond DouaL 

The breadth of the complex, the mighty inversions and imbrications, 
e. g. in Somerset and Devon as far south as Exeter, the violent overthrusting 
of the outer border, the great displacement along certain flaws — even when 
these are posthumous as in the case of the Medina ' fault,' which runs 
across the isle of Wight — and finally the series of granitic intrusions in 
Devonshire as well as in Brittany; all combine to show that we have 
before us the vestiges of a great mountain range. Towards the north this 
range forms a flat arc. The last parts visible in the west, as in the 
neighbourhood of Bantry bay and the coasts of Finist^re, show that the 
remains of this range are continued beneath the Ocean in a west-south- 
westerly direction, and the islands which lie off many of the promontories 
offer a slight indication of this continuation. The eastern parts lie beneath 
the basin of Paris, and here too their strike may be recognized in posthumous 
movements, as in the Pays de Bray. 

This is the great pre-Permiian range of western Europe. The traces of 
its interior an^ presumably most elevated zone lie in Brittany and the 



€H. n] THE VAKISCAN MOUNTAINS 97 

Vend^; for this reason we give these fragments the general name of 
the Armorican chain, 

8. The Variacan mov/ntavns. The rias-coasts of the south of Ireland, 
of Cornwall and Brittany, form a very typical part of the outlines of 
Europe ; corresponding to them, on the other side of the Atlantic, are the 
rias-coasts of Newfoundland and Nova Scotia, which also represent the 
exbremities of a great folded range passing beneath the sea. Starting 
from Ireland, where the contrast between the north and south is so marked, 
we have now examined the horsts which exhibit folding in a common 
direction, as far as the syntaxis between Douai and Valenciennes; these 
combine to form the Armorican arc, the unity of which is shown by the 
corresponding course of the posthumous folds. 

We will now turn aside for a moment from the subject of this chapter, 
that is the description of the Atlantic coasts, and following the same 
method as we have just pursued endeavour to obtain a general view of 
the mountains of Germany, or the greater number of them. This attempt 
could not be made earlier, because it was first necessary to become familiar 
with the idea of a horst, and the closely connected conception of extensive 
subsidences of the surrounding country ; and it was also essential to make 
ourselves acquainted with that peculiar feature — the arcuate course of 
the trend-lines and their syntaxis — which is repeatedly exhibited by so 
large a number of folded mountain chains. 

This digression will be as brief as the difSculty of the subject permits. 
The fragments which I propose to consider are : 

(a) The Devonian TnoimtaiTis of the Rhine, namely the ooUectivQ masses 
of the Ardennes, the Eif el, the Westerwald and the Sauerland, and next 
the Hoehwald, Hunsriick and Taunus. 

(b) The mountains of the Rhine between, Bingen and the lake of 
ConsiaTU^, i. e. the Spessart, Odenwald, Hardt, Yosges and Black Forest. 

(c) The Harz. 

(d) The moun;tain8 of Saxony, namely the Erzgebirge and Fichtelgebirge, 
Frankenwald and Thuringerwald. 

(e) The Sudetes together with the Riesengebirge. 

Many of these fragments are bounded on all sides by longitudinal 
fractures or by oblique transverse fractures ; a part of the mountains of 
Saxony and of the Sudetes sinks beneath the north-German plain; the 
Carpathians approach the Sudetes at right angles to their strike on the 
east and even appear to be driven over them (I, p. 186). 

All these mountain fragments have this feature in common, that they 
experienced a great and general f oldipg towards the close of the Carboni- 
ferous period ; then they were covered by thick deposits of more recent 
sediments and collapsed later, piece by piece, and at divers times ; many 
underwent subsequently a second folding. Thus far, therefore, their 

■OLLAI II H 



98 THE OUTLINES OF THE ATLANTIC OCEAN [part hi 

pasfc history resembles in all points that of the Armorican arc The basal 
beds of the downthrown fragments are not always of Permian age ; in 
some places they are fragments of the higher subdivisions of the Coal- 
measures, pinched into the fractured margin. This is the ease on the 
*north, as on the south margin of the Harz, at Ilefeld, Grillenberg and 
Ballenstedt ^, at Stockheim on the western marginal fracture of the Fichtel- 
gebirge ^, and at Bossitz south of Briinn, on the great fault which separates 
the Sudetes from the Bohemian mass ^ Elsewhere and over vast regions 
the Bothliegende is the lowest member of the subsided series. 

The northern and eastern parts, namely the Ardennes, the Devonian 
mountains of the Rhine, the Harz and the northern part of the mountains 
of Saxony, as well as the eastern and south-eastern part of the Sudetes, 
are characterized by the great development of the Devonian system, which 
is of immense thickness and astonishing uniformity; great r^ons are 
formed by this system alone, and in Moravia and Silesia it presents essen- 
tially the same subdivisions as on the Rhine. The interior parts of the arc 
consist more or less exclusively of pre-Devonian rocks, as in the mountains 
of the Rhine near Bingen, the Erzgebirge, and the western part of the 
Sudetes with the Riesengebirge. 

(a) The Devonian mountains of the Bhine, We are indebted to the 
united efforts of German, Belgian, and French investigators, and in«particular 
to the achievements of H. von Dechen, for our knowledge of the structure 
of this vast region, which extends from Frankfort to DUsseldorf and from 
Mezi^es nearly into the neighbourhood of Paderbom. 

A uniform strike to the north-east predominates in the whole region, 
which for the greater part is thrown into folds overturned to the north, 
or imbricated in true flakes with a south-easterly dip. The coalfields of 
Belgium east of Valenciennes and those on the Ruhr form the outer border. 

The overf olded Coal-measures, from the Boulonnais towards Douai, have 
been recognized as part of the Armorican arc; from the short stretch 
of country between Douai and Valenciennes which marks the region of 
syntaxis, the Coal-measures turn to the east-north-east and north-east, and 
are still affected by overthrusts on a grand scale. Aix stands on an 
anticlinal between two coal basins, the basin of Eschweiler in the south, 
and that of Worms in the north ; but still further to the north, in Holland, 
Goal-measures have been reached by borings beneath the plaint Great 

^ C. E. Weiss, Die Steinkohlen-fuhrenden Schichten bei Ballenstedt am nGrdlichen 
Haizrande ; Jahrb. k. preuss. geol. Landesanst., 1881, pp. 595-603. 

' C. W. Gambel, Geognostische Beschreibung des Fichtelgebirges (Greognostische 
Beschreibuiig des KOnigreiches Bajem, III)^ 8yo, Gotha, 1879, pp. 555-575. 

' A. Makofsj und A. Rzehak, Die geologischen Verh&ltnisse der Umgebung von 
BrQnn ; Verb. nat. Ver. BrQnn, 1884, XXII, p. 82, and geological map. 

* G. Lambert, Nouveau Bassin honiller d^couyert dans le Limbourg bollandais^ 
Rapport, Ann. Soc. g^oL Belg., 1877, IV, pp. 116-130 ; Rapport de M. von Dechen, torn. 



CH. n] THE RHENISH DEVONIAN MOUNTAINS 99 

dislocations here traverse the coal beds, such as the Feldbiss, with its 
contiiiiiation the Miinsteigewand^ which forms the eastern boundary of 
the basins just mentioned K 

On the right side of the Rhine, the continuation of the Coal-measures 
appears to be moved much more towards the north. According to von 
Dechen the Coal-measures of the Ruhr have been recognized near Duisburg 
and Ruhrort; those on the left bank near Yluyn, north of Crefeld^; 
we may therefore conjecture that the Coal-measures which cross the valley 
of the Rhine are either heaved by a flaw, or bent in a sigmoid curve, and 
this supposition is in complete accordance with the lie of the folds which 
follow further to the south. 

Such sigmoid curvature is not rare in folded ranges; the advance 
of one part of the range relatively to another, implied by such a curvature, 
occurs in several places on the outer border of the Alps. The sigmoid 
curve is often intersected by flaws striking across it, and is thus resolved 
into a number of small forward displacements. A ladder-like arrangement 
of the flaw surfaces is then exhibited by the sigmoid. These may be seen 
on the smallest scale, as in thin slices under the microscope ; and on the 
grandest, as in great mountain masses. Studer's geological map of Switzer- 
land shows how the anticlinal of the Molasse, as it proceeds to the east 
from Sankt Gallen and Appemsell, begins to undergo on crossing the Rhine 
a marked deflexion forwards, i.e. to the north-east, and this leads in 
Vorarlberg and the adjacent part of Bavaria to the edge of the Bavarian 
Alps which lie much further to the north. In the same way the great 
Cretaceous limestone mass, which extends from the Toggenburg through 
the Santis to the Iller, describes in the same district, within the Flysch 
zone, a corresponding sigmoid curve. At the same time a part of this 
mass lies in flakes, and flaw planes, striking to the north, cut through the 
Santis (I, p. 116) ». 

The Rhine crosses the folded Coal-measures between Aix and Diisseldorf 
in the same manner as it crosses the Alps above the lake of Constance, 
and it is probable that many of the numerous transverse fractures, along 
which, in the neighbourhood of Aix, von Dechen observed considerable 
displacements both in the horizontal and vertical direction, play the same 
part as the flaws of the Santis. This is the more worthy of remark 

di^ pp. 130-132 ; F. L. Cornet, Notice sur le Bassin hoiuller Limbonigeois, torn, cit., 
pp. 133-142. 

* H. V. Dechen, Geologische and pal&ontologische UeberBiclit der Rheinpiovins and 
der ProTinz Westphalen (Erl&aterangen sar geologischen Kaite, II), 8to, Bonn, 1884, 
pp. 227, 230 ; ako a mach earlier work': Ueber den Zasammenliaag der Sieinkohlen- 
reviere Ton Aachen and an der Ruhr, Zeitschr. f. Beig-, Htltt.-, Salinenw., 1885, III, 

pp. l^o* 

* Id., Uebeisicht der Bheinprovinz, p. 208. 

* e. g. MojdflOTice, Jahrb. k. k. geol. Reichsanst., 1878, XXIII, p. 174. 

H 2 



100 THE OUTLINES OF THE ATLANTIC OCEAN [part ni 

because the earthquakes of Herzogenrath, north of Aix, which oontmned 
for several years, proceeded, according to Von Lasaulx, from the greatest 
of these dislocations, the Feldbiss, and must thus be classed with the group 
of flaw shocks, like so many earthquakes of the Alps ^ 

The most ancient rocks of the Ardennes belong to the Silurian. 
Oosselet distinguishes three Silurian zones, which, however, are not all of 
equal value, namely, the zone of Brabant, of Condroz, and of the Ardennes 
proper \ 

That which is known in Belgium as the zone of Brabant is the 
common foreland of the overthrust region lying in front of both members 
of the syntaxis ; it has been traced by borings beneath the plain, as far 
as Brussels and Ostend. 

Then towards the south follow the overthrust Goal-measures, already 
frequently mentioned, and the zone of Condroz is that ridge which has 
been moved from the south-east over the Coal-measures, recalling by this 
position the anticlinal of the Mendips (I, p. 142, Fig. 17). Folded Devonian 
accompanies this zone, and to the south it is followed by a broad synclinal 
of folded Carboniferous limestone, the basin of Dinant. To the south of 
this broadens out the extensive region of the Ardennes proper, formed 
chiefly of lower Devonian beds. Their strike is directed, precisely like 
that of the Coal-measures, to east-north-east and north-east ; they are like- 
wise overfolded from the south-east, and out of their midst there rise- 
not unlike the gneiss cores of the Alps — four masses which form the 
Cambrian zone of the Ardennes, 

All four masses are built up of Cambrian deposits ; neither upper nor 
lower Silurian is known here. There are two large masses, those of 
Bocroi and Stavelot, the latter called by the Germans the Hohe Venn, and 
two smaller, those of Serpont and of Givonne. 

The mass of Bocroi, that of Serpont, and the Hohe Venn lie in one and 
the same zone of strike; the smaller mass of Givonne is situated above 
Sedan, south of the miEiss of Bocroi. In all the masses, as well as in the 
surrounding Devonian, the beds dip towards the south, corresponding to 
the general inversion. The Bel^an geologists maintain that the relations 
with the Devonian are those of genuine unconformability, but the observa- 
tions of German investigators do not lead to this conclusion, at least not 
with the same degree of certainty. 

The mass of Bocroi lies on the south-west boundary of the Ardennes, 
and is cut through obliquely by this boundary south of Hirson, so that its 
outline somewhat resembles an obliquely intersected ellipse. It is continued 
in any case across the marginal fractures, beneath a part of the Mesozoic 

^ A Ton Lasanlx, Das Eidbeben von Henogenxath am 22. October 1873, 870, Bonn, 
1874, p. 141 et Beq. 
' GK)fl8elet, EaqiuBie g6ologiqa6 dn Noid de la France, fasc. 1, p. 17 et leq. 



CH. n] THE HOHE VENN 101 

sediments of the plain towards the west. Across this mass the Meuse flows, 
disclosing a beautiful section ; Von Trfwaulx has described the constant dip 
towards the south-east, which is exhibited by all the members of this mass, 
from Mezi^res on the south, through the Devonian, the Cambrian mass of 
Rocroi and the Devonian of its northern flank, up to the Carboniferous 
limestone of the basin of Dinant on the north ^. 

The strike of the Coal-measures, directed first to east-north-east then 
to north-east, is clearly repeated in the alinement of the three masses of 
Rocroi, Serpont, and the Hohe Venn ; in the most distant part of the Hohe 
Venn, near Diiren, east of Aix, the strike passes with equal clearness into 
north-north-east. 

The Hohe Venn begins north-west of HouffiJize and proceeds towards 
Diiren. It is broader in its southern part, and is produced with a slight 
curvature first to the north-east and then more and more to the n^th- 
north-east. This is no doubt the same sigmoid as that described by the 
strike of the Coal-measures as they cross the valley of the Rhine. The 
mass of the Hohe Venn is related in the same way to the Coal-measures as 
the limestone mass of the Santis and its continuation to the overthrusting 
of the Molasse. The whole northern part of the anticlinal of the Hohe 
Venn also shows inverted bedding, and the full force of the sigmoid cur- 
vature finds expression on the western side of the extreme end of the mass. 
In this region near M^rode, as Holzapfel has shown, the lowest member of 
the south-east side, the lower Devonian, is completely driven over the 
western limb of the arch, and overlies the upper Devonian and Carboni- 
ferous ; and immediately to the south of this locality several parallel flaws, 
accompanied by as many successive forward thrusts, intersect the western 
limb \ 

^ A. Ton Lasaalx, Ueber die Tektonik and die Eruptiygesteine der franzOsiachen 
Ardeimen, Vortiag in der Herbst-Venammlung des naturluBtoriflchen Vereines der 
prevaaiachen Rheinlande und Westphalen, 7. October 1888, 8yo, Bonn, 1884. Groaaelet'a 
* Faille de Remagne * on the south-east side of the little mass of G-iyonne is a thrast- 
plane. By the same, Sur la faille de Remagne et aur le m^tamorphiame qu'elle 
a prodoit, Ann. Soc. g^ol. Nord, 1883-1884, XI, pp. 176-190. I have not considered it 
neceaaarjr to enter in detail into the disputed question whether an original unconform- 
ability exists or not between the oyerthrust Cambrian anticlinaJs and the lower 
Deyonian, since the imbricated atructare is uniyersally acknowledged. Nor have 
I discuaaed the queation of the existence of traces of a granitic intruaion beneath the 
Hohe Venn ; A. yon Laaaulz, Der Granit unter dem Gambrium des Hohen Venn, Verh. 
natoihiat. Ver. preusa. Rheinl., 1884, XLI, pp. 418-450; Dewalque, Sur lea FUons 
gronitiquea et lea Poudinguea de Lammersdorf, Ann. Soc. g^ol. Belg., 1885, XII, pp. 
158-163. 

* E. Holzapfel, Die Lagerungsyerh&ltnisse des Deyon zwischen Roer- und Vichtthal, 
Teriu naturldst. Ver. preusa. Rheinl., 1883, XL, pp. 397-420, map ; G. Dewalque, Sur la 
tenninaison N.-£. du maaaif cambrien de Stayelot, Ann. Soc. g^oL Belg., 1884, XI, Bull., 
pp. czz-cxzy. In thia reapect my opinion differs eaaentiallj from that of Loaaen, 
a diatingoiahed obaerver . of such phenomena. Loaaen aeea in thia bend of the Hohe 



^ 



i 



102 THE OUTLINES OF THE ATLANTIC OCEAN [part m 

The foregoing will suffice to show to what extent the structure of the 
northern border of the Rhenish mountains resembles the northern border 
of the Alps, and I shall now enter less into detail 

The Cambrian masses of the Ardeimes, the extremity of which we 
have just studied near DUren, are not continued across the Rhine. Many 
series of overthrust anticlinals and synclinals composed of Devonian 
deposits form the mountains far and wide. Their strike is constantly 
directed to the north-east^ and is clearly expressed in the nosing out of 
the folds along the eastern border of the Sauerland, the Westerwald and 
the Taunus, as well as in the general outlines of these mountain fragment& 
Still, towards the south, the overthrusting appears to be of trifling extents 
Synclinal limbs dipping to the north-west, that is, normally inclined, 
alternate more frequently with opposed limbs dipping to the south-east 
On the southernmost edge, from Homburg von der Hohe through Wiesbaden 
up to the valley of the Nahe, Yon Dedien marks a zone of more ancient 
sericite schists and gneisses as the foundation of the folded Devonian 
mountains. Their position is similar to that of the ancient schists of the 
Lizard and Prawle Point, on the inner side of the Devonian f ormationa of 
Cornwall. According to C. Koch, it is along the axis of a symmetrical 
anticlinal, dose to the most southerly edge of the Taunus, that the sericite 
gneiss crops out. It is succeeded on ea<;h side by green sericite schists, and 
these again by the phyllite of the Taunus with beds of quartzite. The 
anticlinal runs to the south-west, crossing the Rhine near Assmanshausen. 
Only its upper members are visible here ; they have closed over the gneiss, 
and further to the south-west on the left side of the Rhine these too 
become concealed beneath the now continuous beds of the Taunus quartzite, 
which in their turn disappear, still further on, beneath the shales of the 
Hunsrlick^ 

Venn, aa well as in that of the basins of the Eifel sitnated further to the south, a torsion 
or deformation of the anticlinal which originally followed the direction of the 
Netherlands, by later Hercynian folding. In that case it seems to me there would not 
be produced a currature of the existing anticlinals, but the formation of new anticlinals 
transverse to the original folds. I may add that the radial system of dykes of 
Andreasberg represented here by Lossen (fig. 12, 1, p. 163), so closely resembles the 
effect of torsion that I too should be inclined to regard this explanation as the moat 
suitable for the present ; on the other hand, I cannot recognize from the data hithexto 
published any turning round of the whole folded system of the Harz ; E. A. Loesen, 
Ueber das AuftretenmetamorphischerGesteinein den alten pal&ozoischen Gebirgskemen 
von den Ardennen bis zum Altvatergebirge, und CLberMen Zusammenhang dieses 
Auffcretens mit der Faltenverbiegung (Torsion), Jahzb. k preuss. geol. Landeaanst., 
1884, pp. 56-112. A. v. Groddeck sees in the Harz only folding in the direction of the 
Netherlands, and this view appears to me to correspond with the observations so 
far published ; op. cit, 1883, p. 73, note. 

^ G. Koch, Ueber die Gliederung der rheinischen Unterdevon-Schichten swiachen 
Taunus und Westerwald ; Jahrb. k preiiss. geol. Landesanst., 1880, p. 198 et seq., section 
on pi. vi. 



CH. n] THE KHINE MOUNTAINS ABOVE BINGEN 103 

(6) T!he mountatTis of the Bhine above Bingen. The southern border of 
the folded DevoniaD region is immediately succeeded by the coalfields of 
the Saar, which present a fundamentally different structure ; the nature 
of the underljring rocks is not known; on the Coal-measures the Roth- 
liegende lies conformably and without any clearly defined limit Gentle 
folds occur and are traversed by numerous fractures. Yon Dechen has 
enumerated many of these in order to show how extreme is the fragmenta- 
tion of the earth's crust in this region \ A mighty fault bounds the Coal- 
measures on the south ; according to Eliver's estimate its throw near 
Bexbad and Sankt Ingbert amounts to 4,000 meters, near Dudweiler to 
8,000 meters. In the south it runs from Saarbriick for some distance past 
Forbach in the direction of Saint- Avoid towards France, probably with 
a bend to the west-south-west, and Lepsius has followed it to the north- 
east as far as Alzey in the Rhine valley ^. 

Thus this transgressive segment is sharply separated from the more 
aootherly heights by a tremendous subsidence which has taken place along 
fiuiltB f oUowing in the main the strike of the Devonian chain, but we must 
not forget that these mighty movements are much more recent than the 
folds, the connexion of which we are seeking to trace. The downthrown 
frs^ment is also gently folded, either by compression during subsidence or 
by posthumous movement. It presents a distant resemblance to the great 
trough subsidence of Edinburgh and Glasgow. 

Bi the north-east the case is certainly different. On the west side of 
the Spessart, east of Hanau and Aschaffenburg, gneiss and crystalline 
schists crop out; the northern outliers of the Odenwald near Darmstadt 
also approach to within 30 kilometers of the southern edge of the Taunus, 
and are only separated from it by the recent deposits of the Rhine valley. 
The Odenwald is followed across the Rhine by the exposures of older 
rocks on the east side of the Haardtwald, emerging, as along the Odenwald 
.and the Spessart, from beneath a thick covering of Bunter sandstone, 
while still further to the south rise the great horsts of the Yosges and 
the Black Forest. 

We have already described the way in which these mountains have 
collapsed (I, pp. 194-197) and the Mesozoic table-lands around them have 
been let down, leaving the existing heights as the ruins of a once con- 
tinuous range. These fragments have been left standing, like the remains 
of an ancient edifice surrounded by the ruins of mighty ramparts; and 
we have now to consider not the circumvallate ramparts but the structure 

' H. TOn Dechen, Uebersicht der Rheinprovinz, pp. 269-275. 

* Op. cit, p. 261 ; on the closely but not exactly corresponding course of a much 
later £uilt, instructive information is given by 6. Mayer, Ueber die Lagerungsverh&lt- 
niflse am Sftdrande des SaarbrUckner Steinkohlengebietes, MittheiL Commiss. f. geoL 
Landes-Unters. ▼. Elsais-Lothr., 1886, 1. 



104 THE OUTLINES OF THE ATLANTIC OCEAN [part in 

of the ancient edifice itself as preserved in the surviving fragments ; we 
must seek to discover the position of the comer-stones, or perchance of 
some remnant of a still recognizable cornice, so that we may recover the 
original outline from the ruins. It would thus be our task to examine 
the direction of the strike and the succession of the rocks in each of these 
several horsts, were it not that this work has already been admirably 
accomplished by Richard Lepsius, who has made use of all the observations 
collected since the time of Peter Merian ; it is only necessary, therefore, to 
recapitulate the results already obtained K 

All the primitive formations on both sides of the Rhine, from the Jura 
mountains to the Taunus, have a common strike, and this is parallel to 
the strike of the Taunus. The whole of this fimdamental mass, now 
broken into fragments, was once folded by a common tangential force 
acting from south-south-east to north-north-west. Gneiss, mica schists, 
and isolated beds of Devonian and Culm follow the common strike, which 
may be disturbed by intrusive granite masses, but only locally or within 
narrow limits. The folding took place towards the end of the Carboni- 
ferous period. On the folded chain, beds 1,200 to 1,500 meters in thick^ 
ness, ranging from the upper Rothliegende to the upper Jurassic, were 
deposited ; then the chain collapsed along great lines of fracture ; and the 
table-land of Trias and Jurassic rocks subsided at least 2,500 meters during 
the interval between the Tertiary period and the present day. 

From the southern end of the Black Forest and the Vosges, as far as 
the overthrust Coal-measures near Aix and on the Ruhr, we thus see one 
and the same system of folds which were formed towards the dose of 
the Carboniferous period. The Black Forest and the Vosges, as far as 
the ancient rocks of the Spessart near Hanau and Aschaffenburg, stand 
in the same relation to the Devonian formations of the Rhine, from the 
Taunus to the Ruhr, as that in which the ancient rocks of Brittany and 
the Cotentin, together with the gneiss reefs of the Eddystone, stand to the 
folds of Devonshire and the overfolded Coal-measures on the north side 
of the Mendips and the Bristol Channel. 

Now for the first time we embrace in one view the whole breadth of 
that great German arc, the outer border of which enters into syntaxis, 
in the north-east of France, with the outer border of the Armorican 
arc It is not inferior in magnitude to the latter, nor are its folds 
less powerful. 

(c) The Harz. This horst presents the form of an ellipse elongated 
from west-north-west to east-south-east, but the strike of the beds is 
transverse, running to east-north-east or north-east. This is very clearly 

^ R. Lepdus, Die oberrheinische Tiefebene und ihre Randgebirge; Fonchnngensur 
deutschen LandeB- and Volkakunde, issued by R. Lehmann, I, Heft 2, 8ro, Stnttgartr 
1885, p. 52 et seq. 



GH. n] THE HARZ 106 

shown in Loesen's beautiful map of the Harz mountains, as well as in 
numerous detailed descriptions of the district. Lossen believes that this 
dcMoainant strike has been disturbed in certain localities by later folding 
in another direction, but Yon Groddeck does not regard this as definitely 
proved \ 

Devonian beds form the greater part of the Harz, and these correspond 
so doeely with those of the Schief ergebiige of the Rhine that Yon Eoenen 
r^ards a subterranean connexion of the two masses as indisputable K 

Two granite masses, the Brocken and the Romberg, rise out of the 
Devonian area ; they do not extend along the strike of the stratified rocks, 
but with this exception resemble in their most important features the 
granite bosses in the Devonian region of Cornwall They produce at 
most only a local deflexion of the general strike. The whole of the 
upper Harz is folded, according to Yon Groddeck, into an anticlinal, both 
limbs of which are thrown into subsidiary synclinals and anticlinala The 
south-east limb is vertical or overturned, while the north-west limb has 
a regular and gentle dip ^. In the central Harz, Eayser describes a great 
anticlinal of lower Devonian, which strikes from Herzberg and Lauterbei^ 
on the south-west border of the mountains to east-north-east or north-east, 
straight across them to Ramberg; it is followed both on the north and 
south by many parallel folds ^. Towards the north, in particular on the 
south-west side of the Brocken, folds occur striking directly on to the 
granite mass ; they are preserved as altered fragments resting on the back 
of this laecolite, which is now exposed by erosion. Most of these folds 
are overthrust to the north-west. In the south-east part of the Harz the 
same strike prevails, but the overthrusting to the north-west does not 
oocur here. On the contrary, it appears from Lossen's description that 
towards the south-east border overfolding of the beds takes place in the 
opposite direction, so that in this part of the Harz fan-structure is met 
with^ In the Schiefergebirge of the Rhine also, the overfolding to the 
north-west diminishes as we approach their south-east border. The Harz, 
indeed, is merely a portion of the Rhenish Schiefergebirge. 

(d) The Tnountains of Saxony. The east-north-east to north-east strike 
of the Black Forest points to the middle of the great fractured area of 
south Grermany. The country to the north of Tubingen and Niimberg 

^ A. von Groddeck, Jahrb. k. preuss. geol. Landesanst., 1883, p. 73, note ; cf. above, 
note 2 on p. 101. 

' e.^. k^ von Eoenen, Ueber Dislokationen westlich und sUdwestlich vom Harz ; op. 
eit., p. 45, 1884. 

' A. von Groddeck, Zur Kenntniss dds Oberharzer Culm ; op. cit., 1882, p. 67. 

* E. KajBsr, Ueber das Spaltensystem am sfidweatlichen Abfalle des Brockenmaasivs, 
insbesondere der Gegend von Sankt Andreasberg ; op. cit., 1882, p. 424. 

* (LoBsen); Erl&uterungen zur geologurchen Specialkarte von Pieussen und der 
ThUnngischen Staaten ; Gradabth. 56, Nr. 30, Blatt Wippra, 1883, p. 36. 



106 THE OUTLINES OF THE ATLANTIC OCEAN [pakp ra 

must have for the base of its foundation the continuation of the Black 
Forest, and as soon as we pass Baireuth and cross the marginal fractureB 
on the east of the great subsidence the older rocks reappear with the strike 
of the Black Forest. 

The chief ridge of the Erzgebirge continues the north-easterly directi<Hi 
as far as the Elbe. On the Bohemian side this is cut off by a great 
fracture, on which a long series of Tertiaiy volcanos once rose. Towards 
Saxony and ThUringia, however, there follows a series of great parallel 
folds driven towards the north-west, which form all the country for a great 
distance into the plain. 

We will first consider the Fichtelgebirge and the Frankenwald. The 
structure of this region is determined by two directions; one, the more 
important, is to the north-east, the other to the north-west. Giimbel, 
however, has justly pointed out that the first or generally dominant 
direction finds expression in folds, the second in fractures ; but it is the 
folds only which now interest us, and not the fractures, which are of much 
more recent date \ From Giimbel's very detailed descriptions it further 
appears that the great granitic intrusions of the Fichtelgebirge and Selb, 
which surround the sources of the river Eger, influence the general strike 
of the gneiss and the ancient schists on the north-east to so small an 
extent that even the Cambrian fragment of Wunsiedel, surrounded on 
three sides by granite, retains the normal strike unchanged. Thus the 
Fichtelgebirge is but the western extremity of the Erzgebirge, and the same 
strike is maintained from the marginal fractures of the south-Qerman 
subsidence, near Goldkronach, as far as the Elbe. The Erzgebirge, however, 
is the principal and characteristic range of the whole mass of Saxony. 

Towards the north this principal range is immediately followed on the 
west by closely folded Palaeozoic slates, and then rising out of them comes 
the gneiss of the Miinchberg in a broad elliptical mass. It extends to the 
north-east from the mai^nal fractures into the immediate neighbourhood 
of the town of Hof , where the Cambrian appears and forms its border. 
The Miinchberg mass presents a most instructive picture of a gneiss core 
of the Alps, denuded almost to its base. The smaller folds and crumplings 
have vanished ; the simplified elliptical outline and the overfolding to the 
north-west remain ^. 

The strike to the north-east is maintained throughout the whole of the 
Frankenwald and also through the Thiiringerwald, so far as it is not con- 
cealed by Permian rocks. We know that the Thiiringerwald is a horst, 
bounded on one side by the continuation of the marginal fractures of the 

^ C. W. Gflmbel, Geognostiache Beschreibung dee Fichtelgebirges mit dem Fnutken- 
walde and dem westiichen Vorlande (Geognostische Beschreibmig dee EOnigreicheB 
Bayern, III), 8vo, Gotha, 1879, in particular pp. 97, 628 et seq. 

' Id., op. cit., p. 635 et paaBun. 



CH. n] THE ERZGEBIRGE 107 

great south-German sunken area, which extends past Cobnrg and Mein- 
ingen, and on the other side by the marginal fractures of the northern 
sunken area in the neighbourhood of Qotha and Weimar. Thus, then, in 
the ThtLringerwald, the beds strike transversely across the mass, and this 
range of heights reproduces the structure of the C!otentin \ The strike of 
the beds in eastern ThUringia, the Carboniferous age of the chief folding, 
and the Permian transgression have been described in detail by liebe ^. 

. Befraining from all detailed description we now pass to the structure 
of the western part of the kingdom of Saxony. This is another fragment 
of the same mass. 

In 1876 H. Credner expressed for the first time the opinion that the 
Erzgebirge is not an independent mountain range, but rather the most 
southerly of three parallel folds which run from south-west to north-east 
through the west of Saxony. The middle fold to the north of it is a very 
symmetrical anticlinal, known as the Mittelgebirge or the granulite moun- 
tains of Saxony, and the most northerly fold, also an anticlinal, though 
covered for the greater part by recent alluvial land, forms the mountains 
of liebfichutz, near Strehla on the Elbe \ 

Just as the elliptical mass of the Mlinchberg gneiss lies in front of the 
western EIrzgebirge, so here, in front of the same chain, lies the ellipse 
of the Saxon Mittelgebirge, directly north of Chemnitz. The towns of 
Glauchau, Bochlitz, Dobeln, and Hainichen may serve to indicate its 
boundary. The ellipse is somewhat larger than that of the Milnchberg, 
its border more regular and its abrasion still further advanced, but the 
main featxu'es of its structure are the same, and Credner's map of the 
range, like GumbeFs map of the Miinchberg mass, gives us the plan of 
a much worn-down mountain core ^. 

' The north-easterly strike of the Thtiringerwald is clearly represented on Richter's 
general map in Zeitschr. deutsch. geol. Ges., 1851, XX, pi. zx; likewise in Credner, 
VersQch einer BUdnngsgeschiclite des Thtlringerwaldes, 8vo, 1855, and in more recent 
woika, e. g. H. Loretz, Beitrag zur geolog^schen Kenntniss der cambrisch-phyllitischen 
Sebiefeneihe in ThOringen, Jabrb. k. preuss. geol. Landesanst., 1881, p. 244, in 
particular the corresponding sheet of the special nmp ; in the present work the 
fractures have been discussed in voL I, p. 193. Lately Loretz has distinguished in the 
south-east of the Thtiringerwald, besides the close and general folding in the direction 
of the Erzgebirge, other wide undulations, on a great or even very great scale, most 
of iriiich also strike to the north-east but some to the north-west ; Loretz, Zeitschr. 
deutsch. geol. Ges., 1886, XXXYIII, pp. 468, 469. 

* K. T. Liebe, Uebersicht fiber den Schichtenaufbau Ostthtlringens ; Abh. geol. 
Speeialkarte, Preuss. etc., 1884, V, pp. 898 and 580, map, et passim. 

' H. Credner, Das vogtl&ndisch-erzgebirgische Erdbeben yom 23. November 1875, 
Zeitsehr. ges. Nat. Halle, 1876, XLVIII, p. 261 ; also by the same, Ueber das erzgebixg- 
isehe Faltensystem, Yortrag gehalten auf dem 2ten deutschen Bergmannstage, 8vo, 
Dresden, 1883. 

* Id., Das sftchsische Granulitgebiiige und seine Umgebung, 8yo, Leipzig, 1884, map, in 
particular p. 61 et seq. 



108 THE OUTLINES OP THE ATLANTIC OCEAN [part in 

A noteworthy interval lies between this great ellipse and the Erzge- 
birge. The Archaean rocks of the Erzgebirge are followed by a Silurian 
zone; then to the east of Hainichen and Frankenberg the gneiss of the 
Erzgebirge appears once more, as a wedge 20 kilometers in length, and 
between this and the border of the ellipse lie Culm and Cambrian phyllite. 
To the south-west, however, in the direction of Chemnitz the steeply 
upturned Culm is covered by upper Carboniferous of the same age as 
Saarbriick beds, and by Rothliegende. These younger beds are horizontal, 
and indicate once more, as in so many other places, the magnitude of the 
disturbances which must have occurred before the conclusion of the 
Carboniferous period. 

It is the most northerly of the three anticlinals of Saxony which lies 
most deeply buried beneath younger deposits. Its oldest rocks crop out» 
as we have seen, near Strehla. Credner has shown that the series of 
exposures of greywacke, which begins at Hainichen and Otterwisch (north- 
east of Boma), runs over the Deditzberg near Grimma to the Collmberge 
and nearly up to the Elbe, corresponds to the southern side of the anti- 
clinal, while the northern flank is visible immediately to the south of 
Leipzig ^. 

This anticlinal reaches the Elbe with an east-north-east strike, but near 
Riesa, not far to the soutli-e€U9t of it, there crop out Archaean rocks, which 
strike from north-west to south-east in the direction of the Biesengebirge 
and the western part of the Sudetes. In these exposures of Riesa we 
may doubtless recognize the continuation of the long gneiss band of 
Grossenhain, which, according to Naumann, proceeds from the district 
north of Dresden, and nms from the south-east in a gentle curve to 
reach the EUbe directly east of Riesa. A broad tract of Palaeozoic rocks 
extends north of this band of gneiss into the neighbourhood of Ortrand, 
and forms the commencement of the great Palaeozoic zone which runs 
from this point through the whole of Lusatia to lower Silesia, past 
Konigsberg, Camenz, and Gorlitz, to the graptolite slates of Lauban; it 
forms an important element in the structure of the Sudetes, and may be 

^ H. Credner, Der Boden der Stadt Leipzig, 8vo, Leipzig, 1853, p. 7. Professor 
Credner has been kind enough to furnish me with the following observations which have 
an important bearing on the questions treated here: (1) Phjllites of Wellerswaldet 
strike N. 75"* £., dip steep to SSE. ; (2) Andalusite mica schists of Clauschwitz, strike 
N. 60"* E., dip vertical; (3) numerous outcrops of lower Silurian greywackes, strike 
N. 45°-50'' E. to N. 60'' E., sometimes N. 70*" E., dip steep to south. The greywacke 
presents in its eastern course towards the Elbe a strike of N. 45'*-50*' £., further to the 
west towards the Collmberge of N. 60°-75*' £. As far as Hainichen, strike WSW.-ESE. 
and dip to the south. SSE. from the Strehla mountains and S. of Riesa some 
isolated bosses of ciystalline schists (gneiss, hornblende schist, &c.) emerge from the 
diluvium. These stiike SE.-NW., that is almost at right angles to the greywacke zone 
of Oschatz-Strehla. The detailed survey of Lusatia is not yet completed. 



n] THE SUDETES 109 

followed much farther still, striking at first east-south-east, then south-east, 
and afterwards more and more to south-south-east. 

Notwithstanding the deflexion of the strike at what is supposed to be 
the north-eastern extremity of the gneiss band of Grossenhain (a deflexion 
compoisated further to the east by a great extension of the greywacke 
zone), I nevertheless believe that the greywacke exposed on the left 
bank of the Elbe up to Oschatz and Strehla must be regarded as a con- 
tinuation of the greywacke on its right bank. In this I agree with 
Namnann and Cotta, who assumed the existence of such a continuity 
many years ago^ 

(e) The Sudetes. A tremendous fracture cuts off the south side of the 
Riesengebiige and extends beyond Dresden. On the outer edge of the 
mountains, as we have already seen, the folded zones on the two sides 
of the Elbe fit together fairly well. The eastern, that is the Sudetic 
side of the mountains, shows this adaptation to a common arc most 
plainly, for towards Moravia the strike becomes almost north-east. From 
this point the several zones of the mountains must bend round through 
90^ in order to adapt themselves in some measure to the arc, and this 
they do. 

I do not propose to discuss in detail the structure of the Sudetic arc, 
since the authors of the geological maps of lower and upper Silesia have 
represented it in so clear a manner. The arcuate plan would be still more 
obvious were it not, apart from the fracture of the inner border, that there 
is a complete absence of correspondence between the outer border of the 
mountains and their structure. From the Katzbach onwards, past Frei- 
bui^ and Jauemig the mountains are cut off obliquely by a long line, 
which cannot be other than a fracture, and beyond Jauemig on the one 
hand and Ziegenhals and Hotzenplotz on the other we see the several 
Palaeozoic zones of the arc with a north-north-west strike disappear one 
after the other beneath the plain. Their continuation probably lies con- 
cealed somewhere to the north of the Zobten. 

The stratified series of these mountains, particularly the Devonian, 
shows, as we have already seen, a close resemblance to that of the other 
fragments of arcs as far as the Rhine. In this region also the most im- 
portant movements occurred towards the close of the Carboniferous period, 
certainly before the Rothliegende. The subsequent folding, not only of 
the Permian but also of the transgressive Cretaceous beds, into synclinals 
which follow the general strike of the arc, as on the Heuscheuer and in 
the basins of Lowenberg and Lahn, shows that posthumous movements 
have occurred similar to those of the south of England. 

' B. Cotta, Erlftnterungen zu Section VI der geognostiBchen Gharfce des EOnigreiches 
Sacbsen, Sto, 1839, p. 46; Naumann und Gotta, Erlftuterungen zu Section X, 1845, 
p. 448. 



110 THE OUTLINES OF THE ATLANTIC OCEAN [pabt ra 

Great masses of granitic rocks rise out of the western part of the chain 
on its inner side ; but Roth very justly remarks that the foldings and 
dislocations of the stratified rocks cannot be ascribed to the granite ; these 
phenomena are more probably connected with movements of the earth's 
crust ' during which opportunity was afforded to the granites to force 
their way upwards/ Beyrich thinks we may assume that these granites, 
like those of the Harz and Devonshire, were intruded during the Devonian 
or the early part of the Carboniferous period \ 

In lower Silesia the chief structural features of the country have been 
determined by F. Boemer, and the researches carried out on Austrian 
territory show that in the neighbourhood of the Altvatergebirge pheno- 
mena of compression occur which equal in intensity those of the Alps, both 
as regards the position of the strata and the alteration of the rocks ^. 

(/) Bwm/mary. We have now reached the eastern extremity of the 
great shattered mountain arc of Central Europe. Its outer border may 
be traced from the syntaxis near Valenciennes as far as the Rhine, and 
after it has crossed the river in a sigmoid curve may be followed some 
distance further along the Ruhr ; its continuation on the east lies in the 
coalfields of lower Silesia and Moravia. Towards the interior foUows 
a broad zone, chiefly Devonian, which extends through the Ardennes, on 
the Rhine up to the southern border of the Taunus, in the Harz, and in 
the more distant parts of the Sudetes. The zones lying still further 
towards the interior consist principally of crystalline rocks; they are 
traversed by more closely folded zones of Silurian, Devonian, emd Culm, 
and form the mountains of the Rhine from the Taunus up to the southern 
end of the Black Forest, the Fichtelgebirge and Erzgebirge with the 
Frankenwald and Thiiringerwald, the Riesengebirge, and a part of the 
Sudetes. 

Like the Armorican mountains this arc was folded chiefly towards the 
close of the Carboniferous period, certainly before the Rothliegende, and 
was broken up at various times. But here too posthumous movements have 
occurred, and these appear most clearly in the Cretaceous basins of the 
Sudetes. Great transgressions over the folds begin in the upper 
Carboniferous, as for example in the coal district of the Saar. 

The highest summits of this ancient range probably stood on the site 
of the Ballons des Yosges, in the south part of the Black Forest, on the 
line running thence to the Erzgebirge, on this range itself, and on the inner 
side of the Sudetic fragment. But nowhere do the outlines of the ancient 

^ Justus Roth, Erl&uterungen zu der geognostischen Earte vom Niederachlesiflchen 
Gebirge und den umliegenden Gegenden, 8vo, Berlin, 1867, p. 390. 

' A new series of researches is now in progress; von Gamerlander, Beisebericht 
auB W. Schlesien, Verb. k. k. geol. Reichsanst, 1886, pp. 294-801 ; F. Becke and 
M. Schuster, op. cit, 1887, pp. 109-119. 



CH. n] THE SYNTAXIS OF CENTRAX EUROPE 111 

monntain cores stand out so clearly as in front of this main line, in the 
gneiss mass of the MUnchberg near Hof , and in the grannlite mountains 
of Saxony. It is thus fitting that the name of this range, which includes 
most of the Oerman horsts, should be borrowed from the land of the 
Varisei or the Vogtland ; and we shall name it therefore the Variscan 
rangej from the Latin name of Hof in Bavaria {Curia Variacorv/m). 

9. The syntaxis of Central Ev/rope. In a previous chapter (I, p. 421) 
we became acquainted with the syntaxes of India. The arcs of the 
Himilaya and those of the Hindu Kush advance like a succession of 
gigantic folds on the surface of two viscous streams moving to the south- 
west and south-east ; they meet each other along a line, indicated in the 
south by the river Jehlam, which may be followed far into the mountain 
region with a direction almost due north, only deviating a little to the east. 
We see, besides, that the folds of the f orelying Tertiary land follow similar 
lines, though the re-entrant angle of the outer border is somewhat more 
obtase than at the junction on the inner side of the chains. 

We must now turn our attention to the syntaxis of the Armorican with 
the Variscan folds. In the great mountain-ranges of Asia the chief obstacle 
to a more exact knowledge of the country lies in its impracticable nature ; 
here the difficulties are of another kind. Everywhere the chain is 
completely worn down, so that scarcely a trace of the original configuration 
of the surface can be recognized ; on the ground plan thus lying before us, 
the true lines of structure, appearing generally as long denuded folds, or 
parallel fractures of great length, must be carefully distinguished from 
a number of others, various in character and of secondary importance, 
or from merely superficial features, frequently regarded as characteristic. 
Id determining the orientation of the so-called mountain systems we 
depend sometimes on the direction of intrusive granite cores or lines of 
volcanos, sometimes on the marginal fractures of horsts, and often even on 
the trend of ridges formed by erosion, or indeed of watersheds. But it is 
evident that> in folded ranges only the folds themselves and the longitudinal 
fractures or thrusts, by which they are accompanied, should be taken into 
account, and it is equally evident that in a profoundly abraded range 
Archaean rocks will form the greater part of the surface; of the sedimentary 
beds which once covered them, only the innermost wedge-shaped extremities 
of the chief synclinals will be preserved. 

From this it follows that the very long pinched-in bands of Silurian^ 
Devonian, Culm, and Carboniferous, which may be met with here and there 
in the Archaean regions, furnish the true trend-lines and guide us to the 
course of the great mountain ranges of former timea 

In the region where we might expect to find the inner branches of the 
t«ro ancient monntain arcs of Central Europe uniting in syntaxis, rises the 
Central Plateau of France. This vast mass, composed chiefly of gneiss and 



112 THE OUTLINES OF THE ATLANTIC OCEAN [pabtot 

granite, terminates in the east, towards the vaUey of the Rhone and the 
Saone, in a long slope running north and south, which, as we shall see 
directly, cuts obliquely across the structural features of the mass, and is 
therefore a fracture. From the neighbourhood of Valence and Privas, 
where a great stream of basalt flows down from the heights into the 
valley, the boundary line of the Central Plateau trends to the south-south- 
west, nearly as far as Le Vigan on the upper H^rault, and even further, to 
Carcassone, if we include the Montague Noire ; thence it runs to the north- 
west, describes a concave arc in the direction of Fig6ac, passes to the east 
of P^gueux, turns round Limoges in a broad curve and then runs east- 
north-east past Saint- Amand towards Avallon at the northern point of the 
Morvan. 

This outline is interrupted by three great bays. Two of them lie close 
together and correspond to the valleys of the Loire and the Allier; one of 
these cuts off the Morvan ; they are separated from one another by the 









'3- III 



1 
I 







spur-shaped range of 
Forez. We have al- 
ready seen that the 
Morvan, which pro- 
jects towards the 
north, is surrounded 
by faults (I, p. 204) ; 
it may be compared 
to the Cotentin and 

I* Fio. 11. l>iaorammatic aeeHon nf the Archaean /auncUUim and the ThUrin^rwald 
fragnuntaof Meacmie rocks on Mont Lo0cre, alter FahreK rm . . « ' 

The two nver vaUeys 
right and left of the chain of Forez are filled with fresh- water deposits of 
the upper Tertiary, and, on the Allier especially, with the products of recent 
volcanos. 

The southern 'bay' differs completely in structure. It opens north- 
west of Montpellier between Le Vigan and Lodfeve and extends with an 
irregular outline to the north-east as far as Mende, to the north-west into 
the region between Lot and Aveyron. This region, surrounded by 
mountains of gneiss and granite, consists of Jurassic limestone forming 
plateaux which for the greater part bear the name of ' les Causses.' The 
plateau of Larzac forms its southern part. The whole region of the 
Causses is generally regarded as a ' Jurassic gulf,' and the Cevennes and 
the Montague Noire as the shores of this gulf K Against this conception, 
however, the steep and bare escarpments of the Causses, as well as the 

^ P. Gonrret, Constitution g^logrique du Larzac et dee GausBes mMdionanx dn 
Langaedoc ; Ann. Sci. g^ol., 1884, XVI, pp. I>229, map and plate. 

' G. Fabre, Snr les preuves de la snbmeraion du Mont Loz^re k T^poque juraasiqae ; 
Bull. Soc. g^l. de Fr., 1873, 3« s^r., I, p. 323. 



[.n] THE CENTRAL PLATEAU 113 



I fanlto by which they are trftveraed, suggest a warning. The 
Cauflsea aie only a downthrown fragment of the vast Mesozoic covering 
which once extended over a great part of the Central Plateau and here 
remains preflerved. The same explanation ^plies to the survival of 
A patch of Eenper, on the summit of the Souquet in the Ceveonea, 
at a height of 1,300 meters ^ 

While the Central Plateau is thus bounded on the north-east, east, and 
flouth by lines of fracture, its extreme western end is seen to dip beneath 
the depression of Poitiers; and since the Archaean foundation ia visible 
in so many places between this western end and the south-easteni part 
of the Archaean hills o£ the Vend^ and Deux-S^vres near Saint-Maixent, 
which lies opposite to it, it is clew that the superficial covering which 
separates the Central Plateau from the mass of Brittany ia of very trifling 
tiiicknees. 

Volcanic masses, such as the Hezen and the eruptive centres between 



Fis. 12. SaclKHt (fcraivA M« Anihracili batin q^CtuklonnM tn Ou loaer IMre, after Ftgte*. 

the Loire and the Allier, the Cantal, Mont Dore and the volcanos near 
Clermont, are superposed on the Central Plateau, but they do not concern 
us here, since our object is to determine the direction of the syntactic folda. 

Let us begin our description in the west. 

An extremely thick band of gneiss and granite runs from the Morbihan 
in a south-easterly direction past Nantes to Fartenay and Saint-Maixent. 
It is clearly the same as that which becomes visible below Poitiers and 
is continued into the Central Plateau. On its northern edge we observe 
a continaons band of Devonian and anthracite-bearing Culm beds, over 
100 kilometers in length, which is pinched into the ancient formations. It 
begins near Nort, north of Nantes, describes like the Armorican arc a gentle 
curve to the north, is cut through obliquely by the Loire near Chalonnes, 

I E. Dnnua, Statittiqne g^ologique, mia^ralof^ne, rndtallurgiqne et pftl^outolo^que 
ia d^partement dn Qaxd, 8vo, Paiia, 1876, p. 155. 
> A. Bnnt, Lee Hooilleres de la Fnuic«, Sro, Patia, 1867, atlas, pi. 23, fig. 1. 



114 THE OUTLINES OP THE ATLANTIC OCEAN [part in 

and reaches the outer border of the horst near Dou^, south-west of Samnar* 
In this long narrow band the plication of the beds is so intense, that the 
same horizon of the Culm, bent back upon itself, actually appears three 
times in the same transverse section, and it was for a long time supposed 
that the Devonian and anthracite were regularly interbedded. In the 
most important exposures the beds converge as they dip steeply down- 
wards. Such is the aspect of the crushed extremity of a synclinal fold 
as it occurs in the Swiss Alps imder the most extreme conditions of 
compression and rolling out. There we meet with Jurassic limestone in 
gneiss, here with Devonian and Culm ; the tectonic process is the same ; 
and thus the band of anthracite on the lower Loire provides us with the 
trend-line of a moimtain chain folded to the north with great intensity, and 
once undoubtedly of great height. 

To the south of the band of granite and gneiss just mentioned, there lies 
near Chantonnay and Vouvant the coalfield of the Vend^ ; it is faulted 
down into the gneiss, and is about 60 kilometers long ; it strikes also to the 
south-east, and, in the midst of the ancient rocks, is accompanied along its 
whole length by an isolated strip of Jurassic sediments. Grand-Eury 
assigns it to the middle Carboniferous. Burat observes that the direction of 
this Carboniferous inlier coincides with that of the south-western border 
of the Central Plateau, which is accompanied by a series of small exposures 
of Carboniferous ; all of them, like the border itself, striking to the south- 
east and subsequently finding their continuation in the coalfield of Aubin ; 
this lies transversely across the region uniting the Montague Noire with 
the Central Plateau, and strikes right into the region of the Causses K 

It would thus appear that the south-western border of the Central 
Plateau must be regarded as one of the Armorican trend-lines, but since 
according to Grand-Eury these exposures are younger than the Carboniferous 
of the Yend^, and the distance between the two groups is also great, this 
question may still be considered as open. 

The presence of Variscan trend-lines in the Central Plateau is very 
clearly marked. The primitive formations may be seen at many places 
to the west of the Vosges beneath the Mesozoic covering, as for example on 
the Moselle near £pinal and at the bottom of many valleys right up to the 
neighbourhood of Bourbonne-les-Bains. To the east of the Morvan the 
granite also crops out at many places near Sombemon, west of Dijon, and 
the exposures indicate a subterranean connexion running from the Vosges 
in a south-westerly direction through the Cdte d'Or towards the Central 
Plateau. 

In the interior of the Central Plateau a long ' line ' appears as the most 
important feature. 

^ Burat, op. cit, p. 199. 



can] THE LINE OP DEOIZE— MAUKIA.C 115 

Near Dedze, on the Loire, a fragment of Cu-bonifen>as crope out in 
the plaiD between tvo faults. It is the foremnner of a long and con- 
tinooiB series of . coalfields, which begin near Sauvigny on the northern 
border of the ancient maes and extend almost without intermption to 
Pl^aoz, south of Mauriac, south of the Dordogne. This line is 160 kilometers 
in length, or if we continue it to Decize, 220 kilometers ; its direction is 
N. 16* E. ; Decize lies a very little further to the east, in accordance with 



PU. 18. JtrattielimMmtpbKlud Ma at gntiu at VMnigBa in Iht MitnOuil, kfter Baltzer >. 

the general mu of the Variacan arc. ' This remarkable arraugemeat,' 
wrote £lie de Beaumont in 1841, ' may be easily explained if we assume 
that the ancient formations have been folded back upon themselves, and 
that these Carboniferoos deposits are only the remains of once much more 
extensive Coal-measures which were overpowered by the fold formed on 

' A. BaltKT, Der mechanuche Contact ron Qneias and Kalk im Bemer Oberlaod i 
Beitr. c geoL Karte der Schweii, XX, 1880, atlaa in iblio, jpl. z, fig. 1. 



116 THE OUTLINES OF THE ATLANTIC OCEAN [part ra 

the surface of the ancient rocks ^.' Only a master mind of deep insight 
could have arrived at this view at that early period ; we now recognize 
a long Yariscan trend-line stretching into the more distant and southerly 
parts of the Central Plateau. 

In studying more closely the Coal-measures of the Central Plateau we 
must now bear in mind that at various parts of its margin, as for instance 
around the spur of the Morvan' as well as towards the south, patches 
of Carboniferous occur steeply upturned along the marginal fractures, 
lying sometimes in little trough subsidences, sometimes at the base of 
downthrown marginal segments, precisely as on the border of the Harz 
and on the transverse fracture of the Fichtelgebirge near Qoldkronach. 
Among these marginal segments we may doubtless include the coalfield 
of Autun on the eastern border, which is partly of Permian age. For 
a knowledge of the inner structure of the mass, however, only those zones 
are of importance which follow definite lines within the mass itself. 

Many years ago, French geologists believed that an intimate connexion 
existed between the older formations of the Voeges and the Central 
Plateau. In 1856 Coquand pointed out that the little granite boss of Serre, 
near Ddle, was a true connecting link between these two great masses, and 

^ ^Saisia dans la ride,'* Dufr^noy et iSlie de Beaumont, Explication de la Carte 
g^ologique de la France, 4to, Paris, II, 1841, p. 625 et seq. £lie de Beaumont expressly 
included this line in his ' Syst^me du Bhin,* but he understood by this term not the 
direction of the folds of the Rhine mountains, but of the Rhine fractures, which cut the 
strike at an acute angle. The curve of the great arc brings the trend-lines of the folds 
themselves into the same direction in the Central Plateau as the fractures present on the 
Rhine. But the truth of the comparison with the infolded synclinals of the Alps 
appears from the fact that in one of the many small bands of the Carboniferous 
formation which lie outside this main line, at Laugeac on the Allier, not only does the 
Carboniferous formation of the border dip beneath the gneiss, as happens often enough, 
but the gneiss lies for a great distance upon the Carboniferous. According to Amiot's 
description the overthrust mass of gneiss has a length of one kilometer, and attains 
a breadth of 500-600 meters ; one shaft traversed 36, another 80 meters of gneiss, 
before it reached the Coal-measures, which however are not inverted. For this reason it 
has been thought that this superposition must be explained by a landslip, but not 
every overthrust shows the inverted, and at the same time the normal succession ; 
Toumaire, Note sur la constitution g^olog^que du d^partement de la Haute-Loire et 
sur les revolutions dont ce pays a ^t^ le th^&tre, Bull. Soc. g^ol. de Fr., 1869, 2* s^r., 
XXVI, p. 1122, and Amiot, Bassin houiller de Langeac, in Etudes des Gites mintoi.ax, 
lOnist^re des Travaux publics, 4to, Paris, 1881, p. 318, atlas in folio, last plate, 
sections no. 1 and no. 10. The tectonic division of the Carboniferous into two series 
has been emphasized by Douville in Compt. Rend, sur les terrains houillers des bords du 
Rhin, LXXIV, 1872, pp. 1323-1325. 

^ T. Ebray, Nullity du syst^me de soul^vement du Morvan, Bull. Soc. geol. de Fr., 
1867, 2® s^r., XXIV, pp. 717-721 ; Michel L^vy, Note sur le terrain houiller des environs 
de Montreuillon, le long de la bordure occidentale du Morvan, op. cit., 1879, 3* ser., 
VTI, pp. 914-919 ; in an essay on the Alpine faults, Ebray says : * C^est dans Fanalyse 
des fuUes que doit se trouver la th^orie de la formation des montagnes,^ op. cit., 1867, 
2« s^r., XXIV, p. 403. 



■ r ■■ ^ 



cH.n] THE VOSGES AND CENTRAL PLATEAU 117 

that a oontmuous line of dislocation might be followed from the southem 
end of the VoBges into the Central Plateau. 

About the same time Foumet observed that the axis of Mont Saint- 
Vincenty an Archaean range which reaches the vaUey of the Sadne with 
a north-easterly direction west of Chalon-sur-Sadne, may be followed 
between the Donbs and the Oignon past Ddle into the neighbourhood 
of Belfort^ and that the disturbances which produced the moimtain folds in 
Sadne-et-Loire probably made their effect felt across the Serre near Ddle 
as far as Qiromagny in the Yosges. Foumet speaks of a chaxTie C^beTmo- 
Voegienne as opposed to the Jura mountains and the Esterel, and even to 
the Black Forest \ 

The direction of this line is south-west to north-east. Between 
Giromagny and Belf ort a fragment of Bothliegende forms the southem 
aid of the Yosges. It is accompanied near Bonchamps by a small exposure 
of Carboniferous, recalling the marginal fragments of the Morvan, the Harz 
and so many other horsts. From this neighbourhood a series of faults 
runs to the south-west, and causes a zone of Trias to crop out for a long 
distance between the Jurassic plateaux. Further to the south-west this 
dislocation or series of dislocations, which, as it happens, forms the 
European watershed over the whole of this distance, coincides with the 
steeper south-eastern border of the granite boss of the Serre ; stiU further 
to the south-west it coincides with a group of great disturbances running 
to the south-west, which for a distance of 60 kilometers intersect obliquely 
the southern prolongation of the Morvan and strike across from the valley 
of the Sadne into that of the Loire. These south-westerly lines bound the 
important pinched-in Carboniferous zones of Blanzy and Creuzot, which 
are separated from one another by a trough subsidence between parallel 
sooth- westerly fractures ; in this the Carboniferous has been let down more 
than 800 meters, as is bhown by borings. These lines reach the vaUey 
of the Loire near Digoin, and in the same direction on the other side of the 
valley another outcrop of workable Coal-measures occurs near Bert ; this 
is let down with the same south-westerly strike, at right angles to the 
direction of the most northerly part of the Forez. Qrand-Eury, it is true, 
has shown that the Coal-measures of Bert are not of Carboniferous, but 
Permian age, and they are consequently not the direct continuation of the 
Coal-measures of Blanzy and Creusot; but the correspondence in position 

' H. Coqnand, M^moire g^ologique Bur TexiBtence du terrain permien et du repr^ 
leotaiit dn Grte Tosgien dans le d^partement de Sadne-et-Loire et dans les montagnes 
de la Serre (Jura), BulL Soc. g^ol. de Fr., 1856, 2» B^r., XIV, pp. ia-47, pi. i, in 
puticnlar p. 40 ; J. Fonmet, De rextension des terrains hoQillers sous lea formations 
aeoondairet et tertiaures de diverses parties de la France, M^m. Ac. Lyon, 1855, V, 
pp. 239 et seq, p. 287 ; also by the same, Apergus sur la stracture du Jura septentrional, 
op. eit., 1861, XI, p. 70 ; see also Jourdy, Orographique du Jura Ddlois, op. dt, 1872, 
2« s6r., XXIX, p. 887 et passim. 



118 THE OUTLINES OF THE ATLANTIC OCEAN [pabtiu 

is 80 dose that we must regard them as the continuation of the Permian 
beds overlying the great Carboniferous band on the right bank of the 
Loire ^ 

The length of this dislocation, from Bert to the southern extremity 
of the Vosges, amounts to about 280 kilometers. 

It is important to note that this line, which may easily be followed 
on the geological map, forms the southern end of the Yosges and the 
marginal fracture of the little horst on the Serre, and then cuts right into 
the mass of the Central Plateau. As may be seen from the brief account 
given above it is mainly of post-Permian age, in its north-eastern part 
even post-Jurassic, but its relations to the folded Carboniferous zones 
of Sadne-et-Loire show that the existing dislocation corresponds to the 
original strike of the chain, and that it is a true longitudinal dislocation, 
like the great fracture of Saint- A void in the region of the Saar. 

£lie de Beaumont designates the southern continuation of the Morvan 
the * chain of the Tarare.' Although it is broken off on the east by the 
great fracture of the Rhone valley, yet it is composed of rocks which strike 
to the south-west, and contains zones of Palaeozoic beds, some of them 
bearing anthracite, which follow the same direction^. The coalfield of 
Saint-£tienne also extends transversely across the chain, and its strike is 
not essentially different. 

It is not necessary to cite further details. The lines Decize — Sauvigny 
— Montaigu — Mauriac and Belfort — Serre — Blanzy — Digoin — Bert, as well 
as the general strike of the zones of the Morvan and the Tarare, show that 
the whole eastern part of the Central Plateau, together with the volcanic 
regions as far down as the Dordogne, belongs to the Variscan arc, and that 
the arcuate ranges towards the middle of the syntaxis show a tendency 
to describe a more acute angle, that is to say a sharper curve, and to pass 
from a south-west to a south-south-west direction^ 

The manner in which the Armorican system enters the Central Plateau 
is not so clearly marked^ but the position of the mountain ranges of the 
Vend^, and their visible continuation below i^oitiers, leave little room for 
doubt that the smaller western part of the Central Plateau belongs to this 
system. 

If we now piece together the fragments thus found, we see that the 
syntaxis of the Armorican and Variscan ranges runs right across France. 
The syntaxis of the overfolded outer borders takes place between Douai 
and Valenciennea Further to the south these ranges must once have 
encountered each other a little east of Paris, their inner branches joining 

^ Grand-Etiry, Flore carbonif^re du d^partement de la Loire, 4to, Paris, 1877, 11, 
p. 511. 
' iSlie de Beaumont, Explication de la Carte geologique, II, p. 152« 



cH.n] THE ALPS AND PYRENEES 119 

at a more acute angle much further to the south, near the head-waters 
of the Dordogne. 

The great ancient ranges have been levelled down, broken up, and sunk 
ia It is only with difficulty that we trace by means of the pinched* 
in ends of the synclinals the position of the great folds by which they 
were built up. But over these fragments in the south-east of England and 
the north of France the ancient folding force reawakened after the 
eoUi^Mse, and produced the posthumous ranges of the Weald, the isle 
of Wight, and the Pays de Bray ; while over the ancient mountains in the 
east were formed the Cretaceous synclinals of the Sudetes. In Central 
Europe another phenomenon occurred. Here, too, more recent movements 
of folding took place, following the old directions. But these did not find 
expression above the ancient fragmentary arcs; they were dammed back 
by the steep fractured margins of the segments left in position or less 
deeply sunk down. Thus the arcuate folds of the Jura, which deviate but 
little from the Variscan direction, run past the southern end of the Voeges, 
where they are checked by the Serre, and beyond the Serre lies Jurassic 
table-land \ In the same way the course of the whole outer border of the 
Alps is predetermined by the position of the f orelying segments. 

The map of Europe shows us even more. The existi/ng contraet between 
the Alps and the Pyrenees ia the aarne as that which once existed between 
the Armorican and the Variscan mountains. 

Our knowledge of the relations of the Alps to the Pyrenees has been 
greatly increased of late years by the efforts of French geologists, and 
particularly by the investigations of Marcel Bertrand in Provence, which 
have rectified many previous views: the most important result is the 
proof obtained by Bertrand that the mass of Hy^res (I, PI. Ill, i ) is not 
a horst serving as a buttress and comparable to the Central Plateau, but 
put of an independent range folded by a force directed towards the north. 

We will now trace the course of the outer folds from. Qeneva to the 
Mediterranean on the map drawn up by Carez and V asseur K 

In the first place it may be observed that according to Bourgeat's 
observations the folds of the Jura mountains between the valley of A in 
and lake Geneva show inversion of the beds seven times in succession, 
possibly owing to the resistance offered by the Serre near Ddle \ Then as 
the curve of the Jura mountains gradually flattens out, the folds assume 
towards the south, in the direction of lake Bourget, an almost due north 
and south strike; but from Chambdry onwards they again turn to the 
soath-south-west, i.e. they again advance towards the plain of the Rhone 

^ Jourdy, torn, cit, p. 876. 

' 6. YaaKur et L. Carez, Carte g^logique de la France, Sheets IX and XIL 
' Booigeat, Snr la repartition des renversemente de terrains dans la r^on da Jnra 
eouprise entre Geneve et Poligny ; Compt. Rend., 1886, CII, pp. 56&-^65. 



120 THE OUTLINES OF THE ATLANTIC OCEAN [part in 

valley ; the cause of this deflexion lies in the interior of the chain where 
the mighty granite mass of the Pelvoux juts out in strong relief. The 
curve in which the folds bend round this mass becomes continually- 
broader; passing from the south-south-westerly direction, they run first 
from north to south, then to south-east, imtil finally ridges appear south of 
the Pelvoux, which strike due east to west, almost perpendicular to the 
main direction of the Alps, and yet they correspond only to the independent 
girdle of the south side of this mass. These transverse ridges, which 
include the Yentoux and the Montagues de la Lure, do not again join the 
outer border of the principal chain of the Alps. A long tongue of Miocene 
deposits, penetrating the range west of Digne, accentuates still more 
sharply the independence of the Pelvoux ; beyond it, forming a new arc, 
for the greater part also convex to the south, begin the outer folds of the 
Maritime Alps, which, closely crowded together, reach the sea near Nice 
with a fresh deflexion of the strike to the south. 

Thus we see first the arc of the Jura flatten out near lake Bourget 
Then the girdle of the Pelvoux advances independently, to terminate 
towards the south in transverse chains, and we are reminded of the fact 
that the Pelvoux represents the extremity of the granitic outer chain 
which proceeds from Mont Blanc. Finally the outer border of the 
Maritime Alps appears, also accompanied by transverse chains, which turn 
first to the south-south-west and then south to the sea. 

This arrangement naturally gives rise to a fairly complicated structure 
in the several chains. Leenhardt and Eilian have recently described the 
transverse chain of Ventoux-Lure ; it is on the whole an anticlinal, cut 
through for the greater part by a longitudinal master fault, by which the 
northern limb is thrown down \ 

The chain of Ventoux-Lure slopes away gently to the south, and is 
followed in this direction, between the Durance and the Rhone, by a new 
transverse chain, the Ldberon ; in the western continuation of this lies the 
chain of the Alpines, which terminates near Aries. 

The approximately east to west strike is continued up to the chain of 
Sainte-Baume, east of Marseilles; which, as Bertrand has shown, is com- 
pletely overfolded to the north for a distance of 15 kilometers. Within 
this chain however, that is, towardei the south, there follows another and 
yet more mighty overthrust, which begins in the neighbourhood of Saint- 
Cyr, and further to the east, near Le Beausset, north-west of Toulon, ia 

^ F. Leenhardt, J^tude g^logique de la region du mont Yentoux, 4to, Montpellier et 
Paris, 1888, 278 pp., map ; W. Eilian, Note pr^liminaire sur la structure g^logique de 
la Montagne de Lure, Basses-Alpes, Gompt. Bend., 1886, Oil, pp. 1407-1409 ; and Note 
g^ologique sur la Chaine de Lure, Feuille des Jeunes Naturalistes, Paris, 1887, XYII» 
pp. 48-55, section. The position of this chain led me to assign it to the girdle of the 
Pelvoux ; Eilian compares it with the more southerly transverse chains. 



cH^n] SYNTAXIS WITH THE HY^RES CHAIN 121 

finally carried so far that a completely cmshed mass of Trias is driven over 
the upper Cretaceous, with a horizontal displacement of at least 6 kilo- 
meters, as estimated by Bertrand ^. We are now in that independent folded 
range, mentioned above, to which belong, according to all appearances, all 
the heights from the mouth of the Rhone to Antibes. 

A very complicated syntaxis thus occurs in the south-east of France ; 
folding movements approach one another, directed on the south border 
of the Pelvoux from north to south, on the south border of the 
Maritime Alps likewise from north to south, and further on, along the 
Var, from east to west, then on the outer border of the chain of Hyferes 
from south to north and on the Var from west to east. This gives rise to 
a general folding of the intervening land with a very variable degree of 
intensity, resulting in the formation of chains of heights, such as the 
Alpines emd the L^beron, in the more open parts of the syntactic region, 
and to an ever closer crowding and crushing of the convergent folds, as the 
border of the Alps and that of the chain of Hy^es approach one another 
from Castellane to the Var. 

Bertrand justly considers the Maures or the chain of Hyferes as an 
important and independent link standing in the space which at present 
separates the Pyrenees from the chain of the Alps. It is evident that the 
Alps and the Pyrenees do not approach each other in a simple V-shaped 
sjmtaxiB. It is clear also that on the south side of the Pelvoux and the 
Maritime Alps the folding force, which elsewhere in the Alps is always 
directed to the north, turns just as completely to the south as it does in the 
extremity of the Carpathians in the south of Transylvania. Perhaps we 
may even succeed in showing that the syntaxis on the lower Var is another 
though minor example of that direct opposition of the tangential move- 
ments, which is displayed on so grand a scale between the Him&laya and 
the Burmese chains on the Brahmaputra (I, p. 452). 

Thus several arcs crowd together in the south of France. The junction 
with the Pyrenees is not to be seen, but the plan of the mountain chains 
lies before us, and we observe the remarkable likeness of the younger to 
the older mountains. 

From this it results that the prevailing tangential movement to the 
m»th, which is characteristic of the Alps and Pyrenees, affected Northern 
Europe, as far as the Caledonian folds extend, before the Devonian, and 
the whole of Central Europe before the Carboniferous. If we consider 
the ancient syntaxis in the Central Plateau of France and its position 
between the Alps and the Pyrenees, we are even justified in speaking of 

* M. Bertrand, CSoupes de la chaine de la Sainte-Beaume, Provence, BolL Soc. g^ol. de 
Ft, 1885, 3* a^r., XIII, pp. 115-150, map ; and Rdle des actions m^caniquee en Provence, 
ttplicaiion de Tanomalie stratigraphique da Beauaset, Gompt. Rend., 13 Juin 1887^ 
CIV, p, 1735. 



122 THE OUTLINES OF THE ATLANTIC OCEAN [pabt in 

the Variscan Alps and the Armarican Pyrevieea : understanding by these 
expressions the older pre-Permian chains. 

In the hor8t8j as we have said, an Mer ErMrope stamds diedoaed 
(I, p. 203). We are now in a position to trace its leading features. 
A great folded system of lofty mountains — proceeding from Centod France, 
out of a region indicated at its northern end by Douai and Valenciennes, 
and at its southern end by the upper course of the Dordc^e — ^runs in are- 
like curves to north-west and west-north-west. This system embraces the 
west and north of France, the south of England and the narrow zone in 
South Wales and Ireland. In Brittany, Cornwall, and Bantry bay, its 
folds may still be seen running out towards the Ocean, beneath which their 
last ruins lie concealed. 

From the same region of syntaxis other great ranges run in arcs to the 
north-north-east and north-east; these comprise all that part of the existing 
continent comprised between the southern end of the Black Forest to the 
coalfields on the Ruhr, the ELarz, the Erzgebirge, and the Sudetes. 

The ancient chains collapse and new terrestrial folds approadi them 
from the south. The Pyrenees and Alps follow the old directions, and 
adapt themselves to the outlines of the ancient fragments still projecting 
which check their course. The Carpathians however, restricted by no 
fractured border, ride over the ancient mountains. 

Thus it happens that in the syntaxis of the Central Plateau of France 
the chief tectonic lines converge towards the south, but in northern and 
central Bohemia, which lie within the arc, towards the north ; so that in 
the former case the characteristic figure is a V, in the latter the same 
inverted A. For the long lines of fracture of the Silurian trough of 
Bohemia, with which Erejci has made us acquainted, and the fault at the 
foot of the Erzgebirge, as well as the line of fracture of the Daubrawa as 
far as Elbe-Teinitz and the faults at the foot of the Riesengebiige, are 
longitudinal lines running approximately in the same direction as the 
fragment of arc which they i^ect. 

Between these arcs, however, there lies, strange and unexplained, the 
mountain fragment of Bioa, striking to the north-west and folded 
according to Oiimbel to the north-east, namely the Bavarian forest with 
the south-western marginal district of Bohemia and opposite to this the 
great gneiss basin of the Waldviertel in lower Austria S with its strike 
directed to the south-south-west. Whether these are the traces of a still 
older sjmtaxis, or what other significance they possess, further investigation 
must decide. 

10. 2%6 Iberian Meeeta. In an earlier passage we distinguished, in 

^ F. Becke, Die kijstalliniBchen Schiefer des niedertaterreichischen Waldvieztels ; 
Sitrongsber. k. Akad. Wiss. Wien, 1881, LXXXIV, p. 546; cf. also the 
of Holger, Gzjzek, Lipoid, and others. 



can] THE IBEBIAN MESETA 123 

aeeordance with the views of MaePherson, three elements in the Iberian 
peninsolA, namely: the Meseta, oompoeed chiefly of ancient rocks; the Betic 
<5ordillera, a folded chain, in the south; ahd the Pyrenees, also a folded chain 
in the north-east. 

The chain of folded mountains which forms the north coast of western 
Africa swerves completely round in Morocco, from east and west to south- 
east and north-west, and finally to south and north, and with this northern 
strike reaches the straits of Gibraltar. At the BAs Torf (cabo Negro) and 
near Ceuta, mica schist and ancient clay slate crop out; further to the west, 
above T^tuan, Trilobites have been found; then follows a zone of red 
sandstone ; west of this the Mespzoic limestone band of the Jebel Musa is 
oontinacd into the Bock of Gibraltar ; stiU further to the west is Flysch, 
which forms cape Spartel. .This mountain arc closes the Mediterranean, turns 
completely round in Spain to the east-north-east, runs in this direction up 
to Alicante, and forms all that part of the country lying south of the Guadal- 
quivir. This is the Betic cordiUera^ traversed by great flaws (I, p. 227). 

Just as the folding of the Alps is checked by the fractured borders of 
tiie opposing horsts, so the Betic folding is arrested along the Guadalquivir, 
and dammed back by the southern border of the Meseta ; this has been 
shown in a convincing manner by MaePherson ^. 

Near Villa do Bispo on the Atlantic coast, a little to the north of cape 
St. Vincent, the southern edge of the Meseta is visible ; Mesozoic beds 
accompany it in the south. It runs fairly parallel to the coast through 
Algarve to the mouth of the Guadiana, accompanied so far by the 
Mesozoic zone; thence it stretches almost in a straight line to the east- 
north-east a long way into the interior, forming the northern slopes of the 
valley of the Guadalquivir and at the same time the southern slope of the 
sierra Morena further to the east (I, p. 294). The way in which the margin 
cuts across the strike of the ancient rocks of the Meseta along the Guadal- 
quivir led MaePherson to recognize in it a mighty fracture; this is the 
fauU of the ChuicUilquivir, 

Calderon's instructive representation of the structure of the Meseta 
and Botella's geological map show the polyhedral outline and peculiar 
eenfigoration of this great horst K 

1 J. MaePherson, Breve noticia acerca la especial estructura de la Peninsula Ib^rica, 
An. Soc. Espan. Hist. Nat., 1879, VIII, pp. 5-26 ; and Uniclinal Structure of the Iberian 
Peainsiila, 8to, Madrid, 1880. For the transverse fracture in the direction of the strike 
on the south border of the Meseta see in particular his little map in the Estudio 
geol6gico 7 petrogr&fico del Norte de la Provincia de Sevilla, Bol. Com. Map. Geol. 
Sspaao., 1879, V, sheet G; for the Betic flaws, 0. Barrois et A. OjBTret, Sur la constitution 
■tiatigTaphique dela chafne B^tique, Compt. Rend., 1886, Gil, p. 1341. 

' 8. Galderon y Arana, Ensayo orog^nico sobre la meseta central de Espa£a, An. Soc. 
IspuL Hist Nat, 1885, XIY, pp. 131-172 ; F. de Botella y de Homos, Mapa geoldgico 
de Espaoa y Portugal, 1 : 2,000,000, fol., Madrid, 1879. 



124 THE OUTLINES OF THE ATLANTIC OCEAN [pabthi 

The fault of the Quadalquivir runs down as we have seen from the 
east-north-east, passes close to the mouth of the Guadiana, and is continued 
through Algarve into the vicinity of cape St. Vincent. The Culm, which 
forms the south-western part of the Meseta in southern Portugal, dips 
towards the Atlantic Ocean beneath recent Tertiary beds, as may be 
seen on the map by Ribeiro and Delgado, but along the coast itself it again 
crops out from under these beds. It is only further to the north, from 
central Alemtejo onwards, that the western outlines of the Meseta are more 
clearly defined. Its rocks, striking to the north-west, advance in sharply 
outlined ridges, which run in rias lines, as it were, into the Portuguese 
plain ; as examples, we may mention in particular the granite ranges of 
Evora. Near Tancos on the Tagus the boundary again becomes rectilinear 
and sharply marked ; it runs thence to the north-north-west, a little to the 
east of Thomar and Coimbra, and reaches the sea not far south of Oporto K 

From this place onwards the rocks of the Meseta form all the iron- 
bound coast of the north-west far across to the Asturias, where the 
Cretaceous mountains of Oviedo, and then of Santander and Bilbao, join 
the Pyrenees with an east to west strike. The north-eastern edge of the 
Meseta, however, is bordered by a broad zone of downthrown Mesozoic 
patches which extend from Cuen9a to the gulf of Valencia and cabo de 
la Nao. 

The structural peculiarities of this great horst are especially apparent 
in the north-west. Fortunately there is a series of excellent treatises on 
this region ; we will only mention those of Wilhelm Schulz and Charles 
Barrels. Barrois in particular has described in a masterly manner the 
chief structural features of GaUqia and Asturias, and the remarks which 
follow are borrowed from his monograph K 

The most ancient rocks crop out in Galicia; these are mica schists, 
green schists and serpentine, as well as gneiss and intrusive granite. They 
strike almost north to south, but with a gentle curve to the west ; in the 
north of Galicia the beds dip to the north-west, in the centre to the west, 
and in the south to the south-west. The dip of the beds is thus directed 
beneath the a/rc and tovHirde the exterior so to speak. On the boundary of 
Asturias, Cambrian deposits begin, broken through in places by granite 
bosses, which play a passive part in the movements of the mountains. 

^ C. Ribeiro e J. F. Neiy Delgado, Carta geologica de Portugal, 1 : 500,000, foL, 
Lisboa, 1876. The small granite maaaes near Cintra do not belong to the Meseta, but, 
according to Ghoffat, are of post-Cenomanian age ; P. Ghoffat, Age da granit de Gintra, 
Jom. Sd. Math. Ac. R. lisb., 1884, XXXIX. 

* G. Schulz, Descripcion geoldgica de la proWncia del Oviedo, Asturias, 4to, Madrid, 
1858, and atlas ; G. Barrels, Recherches sur les terrains anciens des Asturies et de la 
Galicie, M^m. Soc. g^oL Nord, II, lille, 1882, 080 pp. and plate. Among earUer works 
must be especially mentioned A. Pftillette, Recherches sur quelques-unes des roches qui 
constituent la province des Asturies, Bull. Soc. g^ol. de Fr., 1845, 2« s^r., II, pp. 489-482. 



co.nJ THE BASIN OF ASTUMAS 125 

He dip of the Cambrian beda ie preciaely the same as that of the older 
rocks of Galicia, that is beneath the outer arc, and the Archaean rocks are 
consequently driven over the Cambrian. Within the Cambrian, however, 
there follow conformably, with the same dip, a Silurian, a Devonian, and 
finally a Carboniferons arc, always with inverted bedding and increasing 
constriction of the arc, so that on the Rio de Pervia the northern branch of 
the Devonian arc reaches ihe sea with a north-easterly strike, and the 
Carboniferooa is compressed into an ellipse striking from west to east. 
The mi^ty cnrve of a great range is thus visible, and within the curve 
imbricated stmctore. Towards the west, in Qalicia, the most ancient rocks 
form the onter arc ; towards the east, in Astnrias, the Carboniferous beds, 



Fis. li. nt Btuin qfAiluria*, aflM Sohnk and Botalla. 

On the left the ArchMui region (granite, Bneiw, Uld ftncient BchiHtsI ; next to the right, 
Ctmfaruui Knd Siloriui zone d>. ail., with nuBsea of granite j ; then the Devonian zone d, and 
iht infolded Carboniferons limeatone and Calm ca, and the Coal-meanires c^hioh are partly 
MM into the baain, and partlj overlap it transgreMively ; m trauagremiTe Heaozoie patchea ; 
•Bn. lUarial plain of Leon. 

which are the youngest, lie in the middle of the arc, now much narrowed 
by the approximation of its branches. There is a general overthrusting 
towuds the interior. The northern branches strike out towards the bay 
ol Biscay ; the Archaean rocks are directed north to north-north-eaat, the 
Cambrian and Silurian, north-north-eaet, and the Devonian, north-east. 

This great concave curve embraces the whole series of the older rocks, 
ineluding the middle part of the Coal-measnres. It is a remarkable 
(ueonutance, however, that outside these concentric basins the highest 
^vieioQS of the Cofd-measures occur in transgression. These transgresaive 



126 THE OUTLINES OF THE ATLANTIC OCEAN [PARrm 

beds, it is trae, have also been folded subsequently, and in the same 
direction ; Barrois recognizes ' the fact of the recurrence of the same move- 
ments at different epochs ^.' 

In the province of Oviedo there now follows unconformably the whole 
Mesozoic series, together with the Nummulitic limestone, striking to the 
east and folded to the north ; it is regarded as the western part of the 
Pyrenees and does not concern us here. 

It is the outer part of the basin of Asturias, and in particular the 
mountains of Qalicia, which run down to the south-south-east and south- 
east into the great Meseta. A large number of separate works by the 
Geological S\irvey of Spain, now making such admirable progress, show 
us the further course of the rocks in the same direction. In places, it is 
true, patches of Cenomanian rest on the ancient mass, as in Bohemia, while 
the central and eastern part of the Meseta are covered far and wide by 
Tertiary fresh-water beds. Notwithstanding this we observe the following 
facts. The.rsmges running down from Galicia through Tras-os-Montes and 
Beira Alta are formed of ancient folded systems accompanied by great 
exposures of granite. The direction is south-south-east to south-east To 
the south of Salamanca a mighty branch is given off in virgation to the east 
and east-north-east. It consists chiefly of granite and gneiss, and forms the 
sierra de Gredos and sierra de Guadarrama. The other ranges, however, 
continue their course to the south-east, and form a number of almost 
parallel folded ranges of great length, closely packed and in many cases 
slightly convex to the south-west. They consist of Archaean and 
Palaeozoic rocks, including a part of the Carboniferous, as in Asturias. 
The course of this arc may be best indicated by a line drawn from Oporto 
to the south-east past Alcantara, then a little west of C&ceres, past Don 
Benito, then, with a slight bend, through Pozzoblanco to the sierra Morena. 
This line is especially characterized by long granite ridges. Thus with 
a strike directed in general to the south-east, the folded ranges of the 
Meseta reach the Guadalquivir, where they are broken off transversely, as 
we have already seen. 

The bands of granite in this intensely folded mountain sjrstem were 
for the most part intruded, according to MacPherson, during the Carboni- 
ferous period. Here also the upper beds of the Carboniferous lie uncon- 
formably on the folded region, as in Asturias, and the general structure of 
these mountains thus dates, like that of the Armorican and Yariscan 
systems, from «ie last days of the Carboniferous period. 

These folded ranges striking to the south-east are now continued into 
the south of PortugaJ and form the whole south-west of the Meseta. They 
are the same ranges which have already been described as advancing with 
a north-west trend into the Portuguese plain in Alemtejo, like the granite 

^ Barroia, M^m. Soc. geol. Nord, 1882, op. cit., p. 604. 



^■■f- in-.'-i?. 



can] BEND OF THE ASTURIAN BASIN 127 

of Evora. Further south, the strike turns somewhat more to the west- 
north-west, and in the south of Alemtejo, between Beja and Mertola, where, 
according to Delgado, the prevailing dip is to the north-east, imbricated 
strocture would seem to occur. 

In the Spanish province of Huelva lie the famous mines of Bio Tinto ; 
the strike is N. 70*^ W., and thence, according to Delgado, the copper pyrites 
may he followed for a distance of 145 kilometers, to Aljustrel in Portugal, 
the strike gradually turning to N. 50° W. If, however, we include the 
pyrites of Caveira near Grandola, then the length of the band would 
amount to 183 kilometers^. 

These great arcs of the south-western Meseta appear in fact to have 
been folded towards the south-west, and MacPherson's assertion therefore 
applies to them, i.e. that a considerable part of the mountain ranges of the 
Iberian peninsula has not been thrust to the north like the Alps, but to the 
south like the folded ranges of North Africa. 

We now find ourselves confronted by a number of the most difficult 
questions. 

Let us first consider the relation of these arqs to the Betic Cordillera. 

The older folds reach the fault of the Guadalquivir with a south-easterly 
strike; the Betic folds strike to the east-north-east, or almost at right 
angles to them. In precisely the same way, with a transverse strike, the 
Carpathians approach the Sudetes. Here, on the Guadalquivir, however, it 
is quite dear that it is not the direction of the older folds, but that of the 
fractured border, which determines the course of the succeeding younger 
folds. Beyrich arrived at the same result many years ago in the case of 
the Sudetes. 

Let us now glance at the position of the Asturian basins. These basins 
imbricated structure, which dip beneath the outer arcs, indicate in any 
the deflexion of a folded range. It is the outer arcs of the southern 
flank which we have followed from Oporto up to the Guadalquivir, and 
through Alemtejo into the neighbourhood of cape St. Vincent. In just 
the same way the younger folded chain, after crossing from Africa, bends 
round in Gibraltar to form the Betic Cordillera, and if we could make 
a horizontal section, some thousands of feet below the western end of the 
Mediterranean, we should probably find an arrangement of the rocks 
similar to that in the Asturian basins. Between Ceuta and Tangier the 
beds run from south to north, at right angles to the straits of Gibraltar, 
exactly like the rocks in western Galioia, save that here there is more of 
the outer and less of the inner arc to be seen, while in the north only the 
inner arcs are visible. 

Since we know that the mountains of North Africa are only the last 

. * J. F. N. Delgado, Sobre a exietencia do terreno Siluriano no Baixo Alemtejo ; Mem. 
Ac. BeL Lisb., 1876, p. 12, note. 



128 THE OUTLINES OF THE ATLANTIC OCEAN [PAKrm 

of the numerous arcs which mark the southern boundary of Eurasia, and 
that it is these boundary arcs which retain in Europe the curvature and 
folding to the south characteristic of Asia, we obtain the following result: — 
The bending round of the weetemraost mouTUain range of Ev/rasiOj 
where the folding moveTnents pass from a northerly to a soutJierly direction^ 
takes pUice at 'presefnt near OibraUar ; bid towards the close of the Carboni- 
ferous period another gi^eat mountain system had prevumdy arisen, in 
which the change in direction of the folding/ Tnovement from north to sovih 
manifested itself in the place now occupied by the Asturian basi/ns, ue, 
about eight degrees of latitude north of the existing drcumflexure. 

This ancient chain was subsequently worn down and broken up, and the 
new arc was arrested on the Guadalquivir by the fractured border. The 
reconstruction of the continent on the old plan is much more clearly 
manifest here than in Central Europe. The Pyrenees appear to have 
followed the Armorican arc, and the Alps the Variscan, simply as a result 
of the persistence of the southern thrust acting within the space included 
by the residual fragments, but this is not all, and indeed an entirely new 
problem now presents itself. For we have here not simply an ancient 
mountain range followed by a new one, folded in the same direction, but 
an ancient circumflexure followed by a new circumflexure. The region of 
recurvature has been displaced southwards and we do not find, as between 
the Black Forest and the Jura, folds driven to the north-north-west, 
following other folds driven in precisely the same direction, but along the 
fault of the Guadalquivir, the Meseta is formed by that branch of the 
ancient mountains which is folded to the south-west, while opposite to it 
lies the branch of the Betic cordillera, folded to the north-north-weQt 

The question at once suggests itself whether the ancient Iberian range 
is not a fragment of the Armorican arc. In age they correspond ; the rocks 
of Asturias resemble those of Cornwall and Brittany, and both the folded 
regions are certainly continued far beneath the sea; but whether they 
meet or unite, and if so in what manner, it is impossible to determine. 

In th6 Meseta it is again the upper divisions of the Carboniferous 
which correspond with the period of greatest tectonic movement. In the 
last century, when Werner drew for the first time the limit between the 
transition formations' (Uebergangsgebirge) and the 'horizontal forma- 
tions' (Flotzgebirge) he had no other line of demarcation in mind than that 
afforded by the great discordance within the Carboniferous formation in 
Saxony. It is with the same meaning that the term 'Flotzgebirge' 
appears again on Lossen's map of the Harz. Everywhere we find this 
limit. The Coal-measures of Ostrau, which form part of the Sudetes, stand 
in the same relation to those of Bossitz in Moravia, as those of Sama to the 
transgressive Coal-measures of Tineo in Asturias, and in Nova Scotia we 
meet with the same limit on the other side of the Ocean. 



CH.n] PRE-PERMIAN MOUNTAIN RANGES 129 

11. Su/rvey of the pre-Permian rcmgea in Europe, The original con- 
tiniiity of the Armorican horsts, or at least of some of them, was, as we have 
seen, rightly recognized by English and French geologists many years ago. 
A perception of the continuity of a great part of the Yariscan horsts also 
18 evident enough in Lossen's instructive work on the nature of the 
mountain cores which lie between the Ardennes and the Altvater. This 
treatise also shows us how impossible it must have been in the time of 
L. von Buch — ^when observation was not yet sufficiently advanced to 
distingnish lines of fault from the strike of the folds — ^to form any con- 
ception of the structure of the land which would correspond to the existing 
state of our knowledge. A large part of the Armorican horsts have been 
grouped together by Penck also under the name of the ' Central German 
AlpS)' with recognition at the same time of their Carboniferous age \ 

The Yariscan chain makes its appearance first on the northern border 
of the Carpathians and strikes to the north-east and north-north-east in 
Moravia, as though it had here experienced a deflexion similar to that of 
the end of the Carpathians, and as though some trace of the tectonic 
relations of the Iberian mountains were about to recur. Then the range, 
as it forms the Sudetic arc, turns forward to the north, north-north-west, 
and finally north-west, and so passes to the EHbe. 

Next follow with a north-east to east-north-east trend the mountains 
of Saxony, namely the Erzgebirge and Fichtelgebirge, the Saxon folds up 
to T^ipqsig and those folded fragments which are preserved in the ThUringer- 
wald and Frankenwald. These all belong to the inner Yariscan zones. 
13ieir continuation forms the Black Forest, the Yosges, and the exposures 
of older formations in the Odenwald and Spessart; the strike of the folds 
is directed to the north-east, and they proceed to the south-west, past Ddle 
to the Central Plateau of France. Meanwhile the strike gradually passes 
over from north-east and south-west to north-north-east and south-south- 
west, as is clearly shown by the patches of Coal-measures which occur 
pinched into the gneiss as far as the upper Allier. The outer zones of the 
arc are visible, apart from the Sudetes, in the Harz and in the Devonian 
nyyqntft^iTia of the Rhine, from the Taunus up to the Coal-measures in the 
north. The overthrust outer border is only to be seen between Yalenciennes 
and the neighbourhood of Aix-la-Chapelle and in a part of the coalfields on 
the Ruhr. 

The Armorican range rises in the western part of the Central Plateau ; 
in the Yend^, Brittany, and the Cotentin, it trends first west-north-west and 

' K. A Lonen, Ueber das Aufbreten metamoiphischer Gesteine in den alien | 

palftOBoiBchen Gebirgskemen von den Ardennen bis zmn Altvatergebirge, and tlber den i 

Zmammenhaog dieses Aaftietens mit der Faltenverbiegnng (Torsion), Jahrb. k. preuss. | 

geoL Laadesansi, 1884, pp. 56-112 (on torsion of. here note 2 on p. 101) ; A Penck, in i 

Lindetkonde des ErdOieils Europa, published by A Eirchhoff, 1886, 1, p. 818. i 

1 K ■ \ 



130 THE OUTLINES OF THE ATLANTIC OCEAN [PABTin 

then, as it proceeds towards the Ocean, more and more to the west. The 
outer zones are visible in the Devonian reef near Boulogne, in Cornwall 
and Devon, and in the southernmost parts of Wales and Ireland. The over- 
thrust outer border may be traced in a west-north-west direction from 
Douai to Calais. The posthumous folds of the Weald indicate the con- 
tinuation of the range. The overthrust outer border is next again seen on 
the north side of the Mendips; it strikes across the Bristol channel towards 
St. Bride's bay in the south-west of Wales, and thence through the sonih 
of Ireland. The trend turns, towards the Atlantic Ocean, from west-north- 
west to west, and finally, near the sea, almost to west-south-west. 

The Iberian fragment is thrown into folds which trend in a gentle curve 
from Galicia and the north of Portugal to the south-south-east and south- 
east, and with frequent overf olding to the south-west, but they are abruptly 
cut off along the Guadalquivir, opposite the Betic cordillera. They give 
off a branch which forms the sierra de Guadarrama, but in Asturias their 
inner parts bend round so completely that a peculiar basin-like structure is 
produced, with the older members situated on the exterior of the basin, but 
thrust over the younger members of its interior in great flakes. Thus 
the Iberian fragment reaches the sea, striking in Galicia to the north 
and north-north-east and in Asturias to the north-north-east, north-east, and 
finally east-north-east. Its connexion with the Armorican range is not known. 
It thus appears that towards the close of the Carboniferous period great 
mountain ranges arose in Central Europe, which were folded towards the 
north, like the existing Alps. They then collapsed and their fractured 
borders resisted the development of the new folds, which now form the 
Betic cordillera, the Pyrenees, the Alps, and the Carpathians. That 
fragment which still exhibits the ancient syntaxis, the Central Plateau 
of France, now separates the arcs of the Alps from the Pyreneea The 
basin structure of Asturias is repeated in the curvature of the chain near 
Gibraltar. 

In this way Ceniral and Weatern Europe were recomiructed for the 
second time. 

12. European idands. The significance of the islands of the Atlantic 
and of the archipelagos of Europe is appajrent from the facts which have 
already been described. 

The Orkneys and the Shetland idands are parts of the pre-Devonian 
Caledonian range which strikes towards Norway. 

The inner Hebrides represent in part comparatively recent volcanos: 
the outer Hebrides together with the adjacent coast of Scotland correspond 
in position and structure with the Lofoten islands and the east coast of the 
Vest §ord. 

Waigatsch and Nova Zembla are fragments of an independent folded 
range which is syntactic with the northern Ural at Eonstantinov-Eamen, 



can] ISLANDS OF EUROPE 131 

SpUzbergen together with the neighbouring islands up to Franz-Josef 
Land and Bear idand are relics of a plateau with Devonian beds lying 
horJzontaUy, as in Scotland. 

Jan Mayen is of recent volcanic origin. 

The Faeroea are remnants of ancient volcanic flows with intercalations 
of Tertiary plant-bearing beds. 

The great island of Icdand, finally, presents an older volcanic sub- 
stroetnre similar to that of the Faeroes, on which recent volcanos are 
superposed. The substructure contains plant-bearing beds, and is of 
middle Tertiary age. This part corresponds to the similar series of the 
Faeroes and Greenland. The subsequent eruptive formations, however, 
belong to different periods, as is indicated by the distribution of the glacial 
markings and an intervening inundation of the island. The volcanic 
activity must have persisted from the middle Tertiary period down to the 
present day. 

Many investigators have visited Iceland since Eggert Olafssen and 
Bjame Povelsen made their observations on the natural history of the 
island, between the years 1752 and 1757^. But, not to speak of the great 
difficulties due to climate, the impracticability of the country, and its extent, 
these passing visits were the less likely to lead to a clear insight into the 
structure of the island since the investigators as a rule concentrated their 
aistention. on general questions concerning the nature of the volcanic 
phenomena. But now, since an Icelander, Thoroddsen of Reykjavik, has 
devoted himself to the geology of his country and, with as much per- 
severance as ability, is tracing out, year after year, the fundamental 
features of its structure, the facts are beginning to appear with ever- 
increasing clearness. At the same time it becomes more and more apparent 
that the ancient volcanic substructure is traversed like Scania and Spitz- 
bergen by fractures which cut it up into horsts and troughs, and that with 
these fractures the more recent manifestations of volcanic activity are 
connected. Permission has been kindly granted me to borrow from 
Thoroddsen's as yet unpublished observations the following information: — 
Not only is Snaefell's Sysla, the great peninsula half-way down the west 
coast, a true horst bounded on the north and south by lines of fracture, 
and not only. is Faxa Fjordr on the south side, together with the depression 
which continues it into the interior, an area of subsidence — facts which have 
been already mentioned by Eeilhack ^ — but Breidi Fjordr in the north of 

1 Eggeart Olainen og Bjarne Povelflen, Beise igjennem Island, 1752-1757, 2 toIb., 4to, 
Sor«, 1772 (translated into English, 2 vols., 1800-1805). 

' K. Keilhack, Beitr&ge znr Geologie der Insel Island, Zeitschr. deutsch. geol. Oes., 
1886, XXXVIII, pp. 376-449 ; on p. 892 there is a transverse section of the fault trough 
on the south side of the Skard-sheidi, i. e. on the south side of the great horst of Snaefell. 
Hie communications of Herr T. Thoroddsen, of Reykjayik, I owe to the interrention of 



132 THE OUTLINES OF THE ATLANTIC OCEAN [part in 

the peninsula, as well as Hiina Fl<$i on the north coast of Iceland together 
with Steingrines Fjordr, also owe their origin to subsidences, which have 
evidently had the most pronounced effect in determining the contour of tl&e 
table-land. The southern range of Beykjanes is also a horst, on both sides 
of which movements appear to have occurred in quite recent times. 
Nowhere in the south of Iceland do earthquakes occur so frequently as on 
the south side of this peninsula. 

Without anticipating further the publication of Herr Thoroddsen's results, 
it may be pointed out that the discovery of these plateau fractures, and the 
breaking up of the island into horsts, steps, and troughs, while of great 
local interest, suggests questions of a more geijeral kind. In the first place 
we see in the progress of these subsidences the continuation of the pro- 
cesses which have formed the surrounding part of the Ocean. Further it 
must be borne in mind, that all the middle Tertiary rocks of the sub- 
structure contain no oi^nic remains except land plants, and have certainly 
been formed above the sea. A little patch of marine deposits found near 
Halbjarwarstadir, north of Htisavik on the north-east coasts is placed by 
Morch on the horizon of the English Crag ^ : in addition, glacial or post- 
glacial shell beds which attain a height of about 200 feet are met with in 
many places. The island was, therefore, partially inundated at a later 
time. Thus in Iceland the question is very definitely suggested whether 
the movements of the solid land are sufficient to explain the displacements 
of the coast line or whether we must invoke independent changes in the 
level of the sea. 

13. Western Africa. The scant observations hitherto made on the 
north-west coast of Africa leave us free to suppose that Tertiary deposits 
such as we have studied on the west coast of Spain may exist here also. 
From lat. 29** 80^ to 28° N. Duro found the coast formed entirely of beds of 
light-coloured sandstone, undermined by the sea. Similar beds exist in the 
Wady Draa. Scaling the diSa along the coasts we see before us an 
illimitable plain'. The conclusion that the middle Tertiary sea covered 
the western Sahara seems irresistible and the observations of Lenz on his 
journey to Timbuctoo accord well with such a view. 

The adjacent islands are volcanic and some of them bear still active 
cratera An independent series underlies them. L. von Buch distinguishes 
in the Canary islands a foundation of older volcanic rocks on which the 
more recent volcanos were seated. We owe to Doelter the unexpected 

PiofeBBor Nathont of Stockholm ; cf. also T. Thoroddsen, Eine Lavawttote im Innem 
Islands, Peterm. Mittheil., 1885, XXXI, pp. 285, 327, pi. ziv. 

^ 0. A. L. MOrch, On the Mollnsca of the Crag formation of Iceland ; Greol. Mag., 1871, 
VIII, pp. 391-400. 

* G. F. Duro, Exploracidn de una parte de la Costa Nonieste de Africa ; Bol. Soc. 
geogr. Madrid, 1878, IV, pp. 184-199. 




n] 



ISLANDS OFF WEST AFRICA 



133 



discoveiy that Majo, one of the Cape de Yerd islands, is formed to a large 
extent not of volcanic rocks but of slate and limestone, the remnant of an 
ancient continent^. We have already pointed out that marine deposits of 
middle Tertiary age, having the characters of the first Mediterranean stage, 
contribute to the stmctnre of Madeira and Porto Santo, as well as of Santa 
Maria, the most southerly of the Azores (I, p. 288). 

Remote from all these islands, in space as in the nature of their rocks, 
are the five black reefs lying far out to sea in lat O"* 56' N. which are 
designated together as Sunt Paul's rocks. Darwin did not regard them 
as volcanic islands, and Benard has shown, from the specimens brought 
home by the Challenger Expedition, that they consist of peridotite^. 

Aa regards a very large number of the volcanic islands of the eastern 
half of the Atlantic, there is good reason to suppose that the volcanos stand 
on a common foundation. This was the view of L. von Buch and Hartung, 
and has been supported by Calderon in a recent review of the question ^. 

Of this region we may affirm, with much greater certainty than of the 
north, that the visible volcanic islands represent only a small part of 
extensive volcanic regions covered by the sea. The numerous indications 
of submarine activity — such as sudden shocks, eruptions of smoke, or un- 
expected shallows, which have been recorded between long. IS"* and 26^ W., 
especially in the neighbourhood of the equator — led Daussy as early as 1858 
to suspect the existence of a submarine eruptive region situated in about 
lat 0* 22' S and long. 22* W.* 

The leaf-bearing Tertiary beds which appear in the Hebrides, the 
Faeroes, and in Ireland and everjrwhere in the north accompanying the 
basaltic lavas, are not to be found on these islands. To complete the con- 
trast, while the Tertiary beds of the north are never marine, those of the 
islands are exclusively so ; Tertiary beds, however, are restricted to some 
of the islands already referred to. 

Our knowledge of the geological structure of Senegambia and Guinea 
is extremely fragmentary. QUrich's survey of existing observations shows, 
however, that the same structure which it seemed possible to trace (I, p. 898) 
from the far south up to Pungo Andongo (lat 9"" 24' S.) probably prevails 
as far as the lower Faleme and the upper Senegal, as well as over a very 



^ C. Doelter, Die Ynlkane der Cap Yerden und ihre Producte, Svo, Grats, 1882. 

* A. Benard, Description lithologiqne des r^cifs de Saint-Paul ; Ann. Soc. Beige de 
liicioecopie, 1882, 53 pp. 

' D. 8. Galderon, £dad geoI6gica de las Itlas Atldnticas, y su relacidn con los Gon- 
tmentes, Bol. Soc. geogr. Madrid, 1884, IX, pp. 877-899 ; see also Milne Edwards, <V)mpt. 
Bend., 1883, XCVII, p. 1389. 

* P. Daossj, Note snr Tezistence probable d*un volcan sons-marin, situe par environ 
CT 2(K de lat sud et 22^ (K de long, onest, Compt. Bend., 1858, VI, p. 512, reproduced with 
a map showing the distzibation of the observed shocks, in Mallet's fourth Beport upon the 
&cts and theory of Earthquahe Phenomena, Bep. Brit. Ass., 1858, p. 20 et seq. 



134 THE OUTLINES OF THE ATLANTIC OCEAN [PABTin 

large part of the alluvial river basin of the Djoliba [Niger] and the Binne. 
Folded Archaean rocks, fairly various, accompanied here and there by clay 
slate of unknown age, form a basis surmounted by great horizontal masses 
of red sandstone. These are frequently broken up into table mountains 
and do not appear, in the regions studied by GUrich, to exceed a height of 
800 meters. Giirich is inclined to regard the exposures of Foyaite in the 
Los islands, and the olivine gabbro of Freetown (Sierra Leone), as inter- 
bedded with the horizontal sedimentary strata \ 

These characters recur with marvellous uniformity ; we recognize them 
in the descriptions of the banks of the Congo by Leng and Baumann, and 
they are clearly shown in Pechuel-Loesche's geological map of the western 
Congo region. It is near Kalubu, far below Stanley Pool, that the region 
of the horizontal red sandstone begins, which extends into the interior 
beyond the Pool K 

To the north of the equator, however, a new element appears, the zone 
of middle Cretaceous and Tertiary marine beds, which, as it would now 
seem, border almost the whole coast up to the Cunend These were first 
discovered at a few 'widely separated points by Giebel and Lenz (I, p. 398). 
Since then similar Cretaceous beds have been described by Malheiro and 
Choffat in the neighbourhood of Benguela. Here red sandstone, containing 
some gypsum, sulphur, and copper, rests upon gneiss, and is followed by 
Cretaceous beds \ Still further to the south, on the coasts of Mossamedes, 
Anchieta describes the Cretaceous and Tertiary sediments as a selvage, 
hardly 100 meters high and 20-25 kilometers broad, formed of horizontal 
beds which are bounded towards the interior by gneiss ^. 

On the lower Khuiseb, ancient folded schists strike according to Stapff 
north-west and south-east. In Angra Pequefia jointed limestone, some- 
times still in thick beds, rests, according to Schenk, on the horizontal sand- 
stone which covers the Archaean f oimdation ; here great plateau fractures 
occur, which correspond to the course of the coast, and are probably only 

^ G. Gflrich, Beitrftge zur Geologic yon West-Afrika, Zeitschr. dentsch. geoL Ges., 
1887, XXXIX, pp. 96-135; see in particolar A. Pomel, Le Sahara, 8vo, Alger, 1872, 
pp. 28 et seq., and for the coast regions, 0. Lenz, Verh. k. k. geol. Reichsanst., 1878, 
pp. 52, 119, 148, 168, et passim ; also Geologische Karte von West-Afrika, Peterm. 
Mittheil., 1882, pi. i ; G. W. GOmbel, Beitr&ge zur Geologie der GoldkOste in West- 
Afrika, Sitzber. k. bajr. Akad. 1882, pp. 170-196; Ghaper, Note sur la g^logie de 
la possession fran9aiBe d'Assinie, Bull. Soc. g^l. de Fr., 1885-1886, 3* s^r., XIY, 
pp. 105-112. 

* PechuSl-Loesche, Zur Geologie des westlichen Gongogebietes ; Deutsche Rundschau 
f. Geogr. und Stat., published by F. Umlauft, Vienna, 1886, VIII, pp. 289-298, map. In 
the region of the coast asphalt is said to occur. 

' P. Ghoffat, Note pr^Hminaire sur des fossiles recueillis par M. Louren^o Malheiro 
dans la province d'Ai^^ola ; Bull. Soc. g^ol. de Fr., 1887, 8« s^r., XV, pp. 154-157. 

^ J. de Anchieta, Tra^os geologicos da Africa occidental portugueza ; Bol. Soo. geogr. 
Lisboa, 1885, 5» ser., no. 9, pp. 525-529. 



can] CENTRAL AND SOUTH AMERICA 135 

the continuation of the remarkable plateau fractures of Cape Colony, 
previously described ^. 

AU these observations, though scattered over large areas, confirm the 
view that Africa, with the exception of its north-western part, is an ancient 
table-land, and the presence of the middle Cretaceous transgression adds 
another remarkable feature to its resemblance with other table-lands and 
hoists situated far away from it. 

14. The East of Central and Sovih America. From the descriptions 
given in previous chapters of the tabular structure of Florida and the 
resemblance between the cordillera of the Antilles and the borders of the 
western Mediterranean or of the Hungarian plain, we may deduce 
the structure of the coast of Central America. 

A great variety of marine beds of Tertiary or still more recent age 
occurs in this region, and. the classing together of calcareous deposits of 
different age has led to many erroneous conceptions with regard to eleva- 
tions of the land, said to have occurred in quite recent times. We may 
take this opportunity to recall the European characters presented by the 
Meeozoic and Tertiary marine faunas of this part of America (I, p. 522). 

As the subject is of some importance in connexion with one of the follow- 
ing chapters we will now illustrate the structure and succession of strata 
in the outer Antilles by means of an example, selecting for this purpose 
the island of Antigua and making use of information drawn from the 
deecriptionsgiCbrPurves^. 

Antigua has the form of a triangle; its longest side is tu^ed to the 
north-east, facing the Ocean, and measures 25 kilometers. On this side 
the submarine slope is very gentle ; on this side also the most recent beds 
occur, which are continued into Barbados, only 48 kilometers distant and 
separated by no greater depth than 80-40 fathoms. To the south-west, on 
the other hand, the fall is very rapid, and towards the island of Guadeloupe, 
which is not further away than Barbados, we soon meet a depth of 300 
fathoms. The south-western part of Antigua is mountainous and attains 
a height of 1,400 feet; the north-eastern part is hilly, and between 800 to 
500 feet in height ; between these two regions a low plain stretches from 
north-west to south-east, its strata striking in the same direction. 

The more elevated south-western part consists of ancient eruptive rocks 
which are described as porphyrites and volcanic agglomerates (Fig. 15). 
Stratified tuff rests upon these towards the north-east (T^). The tuff is 

1 F. M. Stapff, Earte des unieren Ehtdsebthales, Peterm. Mitiheil., 1887, pp. 202-214, 
map ; A. Schesk, Ueber die geologischen Verhftltniase von Angra Pequefia, Zeiischr. 
devftich. geoL Ges., 1885, XXXYII, pp. 584-536 ; for the south see also A. Moulle, 
Mtooire tor la g^logie g^^rale et snr lea mines de diamants de TAfriqne da Snd, Ann. 
Mines, 1885, 8* a^r., VII, pp. 193-348. 

* J. C. Fbives, Esqoisse g^ologiqne de llle d'Antigoa ; Bull. Mas. R. Hist nat. Belg., 
1885, in, pp. 269-818, pi. xiv. 



136 THE OUTLINES OF THE ATLANTIC OCEAN [pabtiii 

followed by ijie lower or muine siliceooa limestione, which oontoioa 
numerous remains of shells and in addition Prionaetraea divers^ttrmis, 
Solenaetraea taurinengie, Styloooenia lobato-rotv/ndata, Poritee OoUegniwia, 
which all occur in the Mediterranean deposits of Turin, and Alveopora 
daedaUa, which lives at present in the Bed sea, the Indian Ocean, and the 
Pacific (shown hy the black band a, a. Fig. 15). Thia ailiceouB limestone 
is overlaid by sand and volcanic ejections (G) ; then follows upper siliceous 
limestone ; it contains only fresh-water shells and sUicified wood (the first 



Pia. 16. 37M island iif Antigua, aRer Purves. 

white band 6, 6, Fig. 15). Upon it rests anotiier layer of tuff (TJ and out 
of this a peak of trachydolerite project& A strike fault, which traverses 
the whole island, repeats the outcrop of the beds s, b, and T,. We have 
now crossed the moontaina as well as the central plain, and have reached 
the foot of the hills which occupy the north-eastern part of the island. 
They consist of white or yellowish marl and white limestone ; towards tite 
south-west they end in a steep scarp ; to the north-east the beds nuk 
gradually beneath' the sea, giving rise to numerous reefs (c). Here, accord- 
ing to Purves, the specimens of OrbiUndee MaiUdli from Antigua, described 
by Rupert Jones, were in all probability obtained; they indicate the 
horizon of the white Orbitoidea limestone of Jamaica. 



can] ANTIGUA. THE AMAZON 187 

AH the beds hitherto mentioned are very gently inclined to the north- 
east In the most northerly part of the island they are covered by 
horizontal beds of marl containing a mixture of marine and terrestrial 
shells, nearly all of which belong to existing species (d). The lower marl 
beds are purely marine; in the upper beds the association with marine 
species of such genera as Melampus, Physa, and Planorbis, point to a muddy 
shore. A small exposure of basalt appears on the north coast. 

This example shows how complicated the conditions are. The coralli- 
ferous deposits of the first Mediterranean stage (a, a) rest on older tuff. 
The eruptive activity persisted through all that part of the Tertiary period 
which is represented in the island, and the middle Tertiary Orbitoides 
limestone (c, c) sinks so gradually beneath the sea that its fragments, 
sorronnded as they are by living coral reefs, might easily be mistaken 
tiiemselves for modem coral reefs left dry by the sea. 

The same deposits of the first Mediterranean stage which we have just 
mentioned probably also form the foundation of the Bahamas. The Orbi- 
toides limestone also forms a large part of Florida and runs still further 
north in the valley of the Mississippi. In this table-land the gulf of Mexico 
is let down (I, p. 284) and forms the ' fore-sea,' a trough lying in front of 
the Cordillera of the AntiUea The remains of the deep-sea faima of the 
Padfic, living at present in the depths of the gulf of Mexico, and the inter- 
calation of the lacust^rine Grand Oulf Series above the marine Tertiary of 
the lower Mississippi, show that the separation of the Pacific from the 
Atlantic region must have taken place in comparatively recent times, 
under conditions so complicated and variable that we can at present form 
no dear conception of them. 

Let us now turn our attention to the regions further south. 

Marine deposits of Tertiary age extend far up the valley of the 
Orinoco. 

The map of eastern Guayana, constructs by V^ain from the observa- 
tions of Crevaux, shows a more or less east to west strike of the ancient 
rocks composing this region. The intercalated Palaeozoic beds which form 
the northern part of the basin of the Amazon follow the same direction, 
and the course of this coast from Cayenne to the mouth of the Amazon 
thus cuts directly across the strike. For almost the whole of this distance, 
however, a broad band of recent alluvial deposits appears to lie between 
the edge of the mountains and the sea^. 

Now the basin of the Amazon broadens out. From what we at present 
know of the structure of Bnizil we must conclude that the contour of the 
continent also cuts across the strike of the rocks as far as cape Saint-Boque, 

^ C. y^lain, Esqnisae g^ologique de la Guyane fran^aise et des bassins da Parou et du 
Yari, d*apres lea ezplorationB do Br. Creyauz ; Bull. Soc. geogr. Paris, 1885, 7" s^r., VI, 
pp. 458-495, map. 



"^ 



138 THE OUTLINES OF THE ATLANTIC OCEAN [partih 

but from this promontory onwards the course of the coast, at least as far 
as Paraguay, is determined by that of the mountain range& 

As regards this important region we may add to our previous remarks 
the following observations : — 

Mr. Orville Derby has been good enough to call my attention to the 
way in which the direction of this coast is repeated not only in the coorae 
of the Parahyba but also in that of the Parana and the Sfto Francisoo. 
Indeed any map of Brazil will show that on leaving the watershed these 
two great rivers flow to the north-east and south-west on the same line 
parallel to the coast, and then both turn at a right angle to the sea. Their 
mouths are separated by 25 degrees of latitude. The mountain range, 
however, whidi is marked along the watershed on many maps does 
not exist. 

The relation between these two rivers resembles that between the Indus 
and the Brahmaputra, which also flow in opposite directions along the 
same line of strike and then, deflected at right angles, leave the mountains 
by lateral valleys. The Rhine and Rhone aflbrd a well-known example of 
precisely the same kind in the Alps; and in this connexion the Saint- 
(jotthard is only a secondary watershed within a longitudinal valley. 

The structure of the mountain ranges. which determine the direction of 
the coast and the course of these rivers is known in its main features chiefly 
from the descriptions of Orville A. Derby. These mountains are formed 
of folded Archaean rocks. They unite near the upper waters of the Bio 
Grande. The Serra do Mar runs from the south-west along the coast; the 
Serra da Mantiquira proceeds further to the north-east ; towards the north- 
north-east, east of Sao Francisco, lies the Serra do Espinha^o, and another 
range, the Serra da Canastra already known to Eschwege, runs off to the 
north-west between the Rio Pardo and the Supucahy ; it is continued in 
a manner not precisely determined towards Goyaz. 

Against the west side of the folded Archaean region, extending towards 
the Parana, lies a tabular zone of horizontal beds, in which Devonian and 
Carboniferous fossils have been found. This zone lies at a height of 900- 
1,000 meters, and further north of 700-800 meters. Dykes of diabase 
break through it and project in ridges from its surface. 

On the west of this tabular zone of Devonian and Carboniferous rises 
a scarped outcrop which runs from Uruguay to Minas Geraes ; its eastern 
edge reaches a height of only about 1,000 meters on the Parana, and of 
1,200 meters in Minas Geraes. This is the edge of another tabular zone, 
which consists of Permian or Trias with overlying sheets of melaphyre. 

The eastern affluents of the Parana have cut deep furrows in the 
horizontal beds. The decomposition of the diabases and melaphyres 
produces the best soil for the coffee plantations. 

On the SSo Francisco the structure is similar, but between the folded 



n] SESRA DO MAR 139 

^r^h^ATi chains and the region of horizontal strata a tract of folded 
Silnrian appears ^. 

Two elements must thus be distinguished in the south-east and east of 
Brazil, namely, the folded mountain ranges, which, as we have said, even 
include the Silurian on the Sao Francisco, and the flat-bedded table-lands, 
which begin with the Devonian and are very widely distributed towardiS 
the interior. 

In these pre-Devonian mountain chains the older rocks lie towards the 
east, and the folding movement, as far as it has been at present recognized, 
was directed towards the interior. This observation is of great importance. 
It (usigns to the Serra do Mar vn the contiTteifU of South America a position 
wnMar to that occupied in North America by the AppaUtchiavs. 

I believe, writes Mr. Orville Derby, that the comparison of the 
orographic system of Brazil with the Appalachians is based on a true 
homology, and that the chief difference lies in the age of the most 
important elevation, which in North America affectd the whole Palaeozoic 
series, while the formation of the greater part, if not the whole, of the 
Brazilian was pre-Devonian. 

It follows further that the coast mountains of Brazil occupy the same 
position with regard to the Andes as the Appalachians to the cordilleras of 
the west of North America, and that vn the whole breadth of both contments 
ih$ tangential movement is directed from the Atlantic Ocean towards the 
Pacific. 

South of the La Plata we reach the region which lies within the virga- 
tion of the southern Andes. On the La Plata itself marine deposits of 
Tertiary age are present which extend far into the interior of the land. 
They are the commencement of that very varied series of horizontal 
Tertiary deposits, partly marine and partly continental, which, enclosed 
between the branches of the virgation, form the entire coast of Patagonia. 
They will be discussed in a later chapter. 

As we have already learnt, the Falkland islands are a folded fragment 
of Palaeozoic sediments completely alien to the adjacent continent (I, p. 527). 
The reports of the German polar station in Sovth Georgia show that the 
island is formed of folded day slate K 

^ 0. A. Derby, Contribu98e8 para o estudio da geogzaphia phjsica do Valla do Rio 
Onade, BoL See. geogr. Rio de Janeiro, 1885, 1, no. 4, 30 pp. ; by the same, Geographia 
phynca do Brazil, in d'Abren e A. do Valle-Cabral, Brazil Geog^phico e Historico, vol. I, 
1884, translated into English in the 'Rio News,* 5, 15, 24 December, 1884. A geological 
map of this region, by the same author, appears in E. F. van Delden Lafime, BraziliS en 
Ja?a» Verdag oyer de Eoffiecnltunr in Amerika, Asi^ en Africa ; B^dr. tot de Taal-, Land- 
en Yolkenkunde yan Ned. Indie, 1885, IV, 6, pi. i ; cf. also the present work, I, p. 510, 
note 2. 

' E. Moskliaff und H. Will, Die Insel Slid-Georgien ; Deutsche Geogr. Bl&tter, Bremen, 
IM, vn, pp. 113-151, in particular p. 119, et seq. 



140 THE OUTLINES OP THE ATLANTIC OCEAN [PAKrm 

Tristan da Cunha and Diego Alvarez (Qough) are of volcanic origin. 

16. General survey of the Atlantic coasts. The numerous facts jost 
discussed lead to the conclusion that a certain amount of symmetry exists 
between the two sides of the Ocean. In some cases comparison is im- 
possible, in others a correspondence is striking, even if difficult to explain. 
I will now attempt a representation of the facts, taking each region in 
turn. 

(a) To the north, in the middle of the Ocean, rises the wedge-shaped 
mass of Oreefjfda/nd ; on each side lies the sea. 

(6) The eastern boundary of this sea is formed in the first place by 
a range of ancient gneiss which strikes down from Magero to the jagged 
peaks of the Lofoten, and is visible in the Hebrides further to the south- 
west. — ^The western boundary, in Davis strait and Baffin bay, consists 
likewise of a jagged gneiss range, which comes from the north and follows 
the coast towards cape Walter Bathurst^ through Cumberland and Labrador 
up to the straits of Belle Isle. 

(c) A folded range of pre-Devonian age, the (kdedonian mountaine^ 
beginning probably in Itorway, runs through the Shetlands and the 
Orkneys, Scotland, Wales, and a great part of Ireland. The Scotch horsts 
lie in its strike. — No corresponding range is known in America. 

(d) On the east of the gneiss range and the Caledonian zone there 
follows in Europe the great Baltic shield. The folded Silurian beds of 
Norway appear to pass into the flat-lying Silurian of its western border. 
Archaean formations are exposed in the middle of the shield ; the glint line 
completely surrounds it. The Varanger fjord, the lakes of Lapland, the 
gulf of Finland, lakes Ladoga and Onega and the gulf of Onega mark the 
edge of the shield. The shallow gulf of Bothnia lies upon it. 

A similar structure is repeated in Canada. West of the gneiss range 
of the coast lies the great Canadian shield. It is surrounded by flat-lying 
Palaeozoic beds. The great lakes mark its southern boundary, which runs 
down through lakes Winnipeg and Athabasca, the Slave lake, Marten lake 
and Bear lake to Coronation gulf ; then probably through Simpson strait 
and perhaps through Melville peninsula. Hudson bay, of no great depth, 
lies upon it. 

In Europe, as in North America, there is a shield, a shallow sheet of 
water, a ring of glint lakes, and a flatly bedded Palaeozoic border. 

(e) On the west coast of Ireland where the Caledonian mountains, 
striking to the south-west, disappear, the folded ranges of another chain 
arise, which proceed in an arc from the interior of the continent, striking 
first to west-north-west, and then to the west. It is folded to the north. 
This is the Armorican a/rc. It reaches the south-west coast of Ireland 
with a westerly direction ; this coast, Cornwall and Devon, as well as the 
coasts of the north-west of France, owe their rocky shores to the way in 



can] SYMMETRY OF THE ATLANTIC 141 

whieh these folds strike out to sea. This chain arose for the most part 
bef oie the end of the Carboniferous period. 

In the same way Nova Scotia and Newfou/ndLaiid are detached from 
die contour of America and show us the folds of a great mountain range, 
whieh coming from the south-west assumes gradually a west to east 
direction. This range also is folded to the north like the Armorican arc ; 
it also attained its chief development before the close of the Carboniferous 
period. 

(/) In Europe there next follow the Pyrenees. It is impossible to 
diaoover a corresponding range in America. 

(g) In Europe we now arrive at the jbasin of Asturias. There is nothing 
similar to be met with in America. 

(h) We then come to the Mediterr^ean on the east side of the Ocean. 
It is closed in near Gibraltar by a chain folded towards the exterior and 
bent round in a sharp curve. This is the most westerly prolongation of 
the Betic cordiUera and presents at the same time the only instance in 
which the outer margin of a folded range reaches the east coast of the 
Atlantic, without sinking beneath the Ocean in a rias coast. 

On the west side of the Ocean, though further to the south, we see the 
Oaribbean sea, girdled about by the cordiUera of the AntiUeSr which also is 
folded towards the exterior eaxd sharply curved round : this again is the 
only place on the western side where the outer border of a folded range 
reaches the coast without being faulted down as it strikes out to sea. 

It is not possible to continue these comparisons further to the south. 

Greenland lies symmetrically between two continents. The gneiss* 
range of the Lofoten corresponds to that of Labrador; but the objection 
may be raised that in Europe the former is interrupted for a great distance 
and that the latter in America is little known. The Caledonian range is 
absent in America. The two shields correspond and the two pre-Permian 
rias coasts. The Pyrenees and the basin of Asturias are not represented 
in America; on the other hand there is a striking resemblance between 
the two Mediterraneans with their girdle of folds. 

It may be observed that certain elements are represented in Europe 
twice over as it were; thus the Armorican rias coast is repeated by the 
I^^renees and the basin of Asturias by the girdle of the Mediterranean. 
This duplication, however, arises from the repetition of the pre-Permian by 
Uie Tertiary ranges, the building and rebuilding of Europe revealed by the 
horsts. 

The Caledonian folds do not come into question here; they are not 
known in North America, and in the northernmost part of Norway no 
Caledonian folds are interposed in the course of the horizontally stratified 
masses of the lake region as they extend to the west, up to the dislocation 
which separates them from the gneiss of the Lofoten. 



142 THE OUTLINES OF THE ATLANTIC OCEAN [partih 

Dingle bay on the west of Ireland, on the outer border of the Armorican 
are, corresponds consequently to Belle Isle straits between Labrador and 
Newfoundland. Both lie approximately in the same geographical latitude. 
But thence to the south the homologous parts of Europe are much more 
closely crowded together than their Mexican equivalents, and thus the 
Mediterranean lies considerably further north than the Caribbean sea- 
It is to the repeated reconstruction of Europe that the peculiar diffi- 
culties are due which have always confronted tectonic studies in this 
continent. Elsewhere the mountains are more uniform in structure, planned 
on broader and simpler lines. Even if in many cases we have not been 
able to follow the bordering faults all round the mountain fragments which 
we have represented as horsts, even if we prefer to assume that the 
Central Plateau and Brittany, for example, have been separated across 
Poitiers not by subsidence, but by erosion, yet this will not alter the result 
From the lower Guadalquivir up to the neighbourhood of Briinn the 
younger folded ranges are faced by the walls of faults. What is there 
known as the sierra Morena is here called the Manharts-Gebirge, and the 
only question which remains is to what extent the fragments to the north 
of these walls are to be regarded as independent of one another. This 
question, however, is only of secondary importance, since within the 
mountain segments themselves mighty dislocations are present of the most 
various age, as for example the marginal fractures of the Scotch trough, 
the fault of Saint-Ingbert in the basin of the Saar and the Letten kluft 
near Przibram. 



143 



CHAPTER III 
THE BORDEKS OP THE PACIFIC OCEAN 

Ne^ Zealand. Australia. New Caledonia. The sea of Banda, Borneo. Cochin-cliina, 

Tongking. The Philippines. Formosa and the Liu-Kiu islands. Japan. The Knriles ' 

and Kamchatka. General sorvej of the island arcs. Eastern China. North-eastern ^ 

Asia. The arc of the Aleutian islands. The west coast of America. i 



1. New Zealand, Since the brilliant description of this island group by 
F. von Hochstetter^ onr knowledge of its structure has been greatly 
increased by the contributions of Hutton and Julius von Haast and by 
the work of the Geological Survey under the direction of J. Hector. 
Space precludes more than a summary of the most important results 
achieved by these laborious investigations; I am fortunate, however, in 
having at my disposal, not only a large body of already published material 
and the general geological map by Hector, but also information afforded 
me in friendly communications by Herr von Haast, and a detailed manu- 
script sketch which I owe to the kindness of Captain Hutton. 

The series of marine beds in New Zealand is very complete. The 
Silurian is represented by at least two series of f ossilif erous strata, a lower 
group with Graptolites and an upper group with Trilobites ; the Devonian 
is less distinct; the Carboniferous limestone contains Spirifer himLcaiua^ 
Productus hrachythaeTU8y and other characteristic species; above it lies 
the group distinguished by Glossopteris, but the marine Carboniferous 
beds, which in Australia overlie the coal with Glossopteris, are not recorded 
from New Zealand. The Trias is represented by the Wairoa beds with 
Pteudomonotis and Halobia. Then follow beds distinguished by Ammo- 
nites and Brachiopods, which are assigned to the lias or lower Jurassic 
(Gatlin's River amd Bastion Series), and plant-bearing beds with Macro- 
taeniopteris lata, which are correlated by Hector with the Bdjmah&l series 
of India. This flora, corresponding approximately to the middle of the 
Qondwana series, is evidently the same as that which extends with such 
an astonishingly wide distribution through almost the whole of Eurasia ; 
as regards its relations to the marine deposits of New Zealand, there is at 
present great dearth of information. Deposits with Bdemnites] AuatraXis 
are rightly or wrongly regarded as representatives of the lower Creta- 
ceous; then follow very f ossilif erous sediments of the middle Cretaceous 
and a series of marine Tertiary formations. The sediments are accompanied 



144 THE BORDEES OF THE PACIFIC OCEAN [pabtih 

by various volcanic rocks, which begin in the Palaeozoic aera and, on North 
island, are still in process of extrusion. 

Varied as is the succession of the strata, the structure and features of 
the land are no less so. In the south it forms a true Alpine r^on where 
several peaks rise above 3,000 meters ; in the middle of North island lies 
one of the most remarkable volcanic areas in the world. The outlines of 
the islands, however, correspond only in part with the trend of the 
mountain folds; fracture and subsidence have had a large share in their 
formation. 

Hochstetter had already suspected that Cook strait and Foveaux strait, 
which separate the three islands, were due to the subsidence of mountain 
blocks ; he was also aware that the mountain chain which follows the east 
coast of North island from East cape to Wellington is continued on the 
other side of Cook strait between the east coast of South island and 
the river Awatere, and that this continuation lies to the east, outside the 
trend of the principal chain ^. More recent investigations in the south, 
however, have revealed the remarkable fact that in the southern half of 
South island two directions of strike and folding encounter one another 
almost at right angles, and the knowledge we have acquired in other parts 
of the earth leads us to conclude that in this region two iikdepend^nt 
unilateral chavas meet in ayntaada. One of these chains trends to the 
north-east, and its oldest rocks lie to the north-west and west; all the 
mountain fragments of North island belong to it. The second chain runs, 
so far as it is known, to the south-east, and its oldest rocks crop out on its 
south-west side} it includes the southern part of South island together 
with Stewart island. On the east coast near Dunedin it is broken off 
transversely across its whole breadth \ 

The chief structural features which are developed from this plan are 
as follows : — 

A long narrow band of gneiss and ancient granite follows the west 
coast of the middle part of South island. It has been described by 
Herbert Cox \ Only a few Palaeozoic patches lie along the coast on the 
western flank of these ancient rocks ; towards the east, on the other hand, 
they furnish the base of an extremely thick zone of slates, also Palaeozoic, 

^ F. yon Hochstetter, Geologie von Neu-Seeland ; Reise der Osterreichischen Fiegatle 
Novaia um die Erde, Geologischer Theil, I, 4to, Wien, 1864, p. xlvi, 2, at paasim. 
Hochstetter'B first impression of Cook strait was not that of simple fracture, but of 
a displacement of the islands in an opposite direction by a mighty lateral force ; bj the 
same, Lecture on the Geology of the Province of Nelson, New Zealand Government 
Gazette, Nelson, Oct. 22, 1859, p. 101. 

* 1 refer the reader to the diagram of South island in Button's Sketch of the Geology 
of New Zealand ; Quart. Joum. Geol. Soc, 1885, XLI, p. 195. 

' S. H. Cox, Report on Westland District ; Rep. Geol. Surv. New Zealand, 1874-1876, 
Wellington, 1877, pp. 63-93, map. 






I 



cam] NEW ZEALAND 146 

which f<nm the highest summite of the New Zealand Alps. Here is 
otnated mount Cook, which attains a height vtuiouBly estimated at from 
3,762 to 8,963 meters ; great glaciers descend from this lofty range ; a vivid 
aod instmctive picture of the region has been given by Von Lendenfeld *. 
Towards the east the range slopes down and is followed by a long synclinal 



Fia. 16. New Ztaiand. 
1, Arcbaaui ; p, DrystalUne nluBta and PaUeozoic ; Irj, Triu and Jutbbuc ; tri, Cretaceoos 
ud Ttrttuy ; dotted araM, volcanic regions, chiefly acid Uva« and tli«ii asaociated tu^ ; 
bUtk, principal volcauoe and basic lavas ; white, recent alluTlDin. 

of Hesozoic beds, till Uie Palaeozoic foimdation again rises in an anticlioal ; 
then comes a broad plain on the east which onit^ the border of the moun- 
taans with a block -like mountain mass projecting far out to sea. This is 
' R. V. Lendenfeld, Qer Taunan-Qletaoher mid wine Umgebang; Petenn. MitUi., 
ErgtoEiuigdkefl, Nr. 75, 1884, 80 pp., mapa. 



146 THE BORDERS OF THE PACIFIC OCEAN [pabtih 

Banks peninsula. Haast, who has examined and described the whole of 
that part of the island just referred to, has shown that on the west mde 
of this mass some ancient schists crop out, and associated with these quarts- 
bearing porphyry, but all the rest of the mass, which attains a height of 
927 meters, rising from an area as large as the base of Aetna, oonsistB 
of a number of closely crowded craters of various age, standing beside and 
above one another. Some of these reveal the same radial arrangement 
of the dykes as distinguishes mount Venda in the Euganean r^on 

(I, p. 146) K 

The ancient rocks of the west strike to the north-east, and pass into 
elongated mountain cores ; one of these reaches the sea on the west coast 
of Tasman bay. On North island neither gneiss nor granite is exposed. 

Palaeozoic rocks also appear to the west of the crystalline schists of 
Tasman bay ; but the principal zone, which runs from mount Cook to the 
north-east and includes many lofty peaks, reaches Cook strsdt, passing 
mount Franklin on the east side of Tasman bay ; it then breaks up into 
a number of peninsulas, islands, and cliffs, and sinks beneath the sea 
between dTTrville island and Blenheim. To the east the principal Palaeo- 
zoic zone is followed by the Mesozoic synclinal, which still further east, 
beyond the river Awatere, is followed by Palaeozoic anticlinals; towards 
the east coast these in their turn are accompanied by a Mesozoic zone. 

These anticlinals of Palaeozoic rocks lying east of the Awatere form the 
Eaikoura chains, which attain a height of over 8,000 feet; Hochstetter 
rightly recognized them as the continuation of the long Palaeozoic range 
of North island : this begins near WeUington, and as the Tararua chain, the 
Buahine chain, or under other names attains in several places a height of 
over 5,000 feet ; it finally reaches the north coast east of the bay of Plenty. 
A Mesozoic zone and subsidiary Palaeozoic chains border its east side 
from cape Palliser to cape Runaway, so that the course of this part of 
the coast corresponds with the actual strike of the bed& 

The long and narrow Palaeozoic band running from Wellington to 
the north coast, which for brevity we will call the Buahine chain, is 
the only continuous folded range of North island. Towards the west 
it is followed by the extensive volcanic region of lake Taupo with the 
gigantic volcanos Tongariro and Ruapehu. On the west side of the island 
another volcano, mount Egmont, emerges in a great circular outline from 
the sea. Tertiary and even more recent sediments form the north coast 
of Cook strait between this broad cone and the south end of the Ruahine 
range. To the north-west Palaeozoic rocks are known in many localities ; 
up to the North cape, however, and beyond it, these are only the isolated 
fragments of the sunken range, between which recent volcanos appear in 

^ Julius V. Haast, Geology of the Provinces of Gantexburjr and Westland, 8to, Christ- 
church, 1879, pp. 824-349. 



m] NEW ZEALAND 147 

many places, especially near Auckland. More recent sediments, tuffs, and 
lavaa unite these fragments by a superficial covering, and so form the 
greater part of North island. The regular arc-like curve of the north- 
western shore is formed by littoral accumulations which run from one 
mountain fragment to another. The north- western coast, therefore, in no 
way represents the actual trend of the mountains. The south-east coast, 
interrupted by Hauke bay, alone corresponds with this direction. 

The principal chains of South island disappear at Cook strait, and it is 
only a secondary chain which is continued across North island as the 
Buahine range. We naturally look for the continuation of the principal 
chains of the south beneath the volcanic region of the north. ' Probably,' 
says Hutton, ' the ge-antidinal of the South island runs through the centre 
of the North island from Wanganui to the bay of Plenty \' Hochstetter 
has already distinguished under the name of the Taupo mountains a series 
of recent volcanos running right through North island parallel to the 
Buahine range and to the west of it. This zone, characterized by acid 
lavas, runs in a north-easterly direction from the mouth of the Wanganui 
on the coast of Cook strait to the island of Whakari (White island), an 
active volcano which rises from the bay of Plenty. Ruapehu (2,793 
meters), Tongariro (2,582 meters), lake Taupo, Tauhara and Putanaki 
(mount Edgecumbe) 'lie on this line, which Hochstetter regarded as the 
boundaiy or margin of a downthrown area'. 
Let us return to South island. 

The syntaxis does not take place at an acute angle, but in a fairly open 
carve. Possibly the form of syntaxis aa projected on the map is determined 
by the degree to which denudation has advanced. 

A Palaeozoic spur runs out from the region of mount Cook to the south- 
east, and approaches the mouth of the Waitaki on the east coast in a gentle 
curve. A second arc lies to the south of this river. Finally, the whole of 
the Palaeozoic zone, here very broad, swerves round near lake Wanaka, in 
west OtagOy from south-west through south to south-east, and then ends 
by breaking off on the coast near Dunedin as we have already seen. On 
the south-east coast this zone extends to Molyneux bay, and is there 
foUowed on the south by a very remarkable Mesozoic zone, which includes 
the Hokanui mountains and is distinguished by a particularly rich suc- 
cession of Jurassic beds, both marine and continental, as the observations 
of Cox and M'Eay have shown \ This zone takes part in the folding, 

* Hatton, Sketch of the Geology of New Zealand, p. 197. 

* F. von Hochstetter, Geologie yon Nea-Seeland, p. 92 et seq. 

' S. H. Cox, Report on the Geology of the Hokanai Ranges, Southland, Rep. Geol. 
Snry. New Zealand, 1877-1878, Wellington, 1878, pp. 25-48, map ; Alexander M*Eay, 
Notes on the Sections and Collections of Fossils obtained in the Hokanui District, torn, 
cii, pp. 49-90 ; by the same, Mataura Plant-beds, Southland County, op. cit., 1879-1880, 

L 2 



148 THE BORDERS OF THE PACIFIC OCEAN [PABrm 

strikes to the south-east, and finally almost due east. To the south and 
south-west it is again followed by slates, forming the north part of Stewart 
island, and then finally by gneiss. 

The extreme south-west of South island from Milford sound onwards 
consists of gneiss; the coast is very steep and cut into by deep fjords; 
gneiss also forms the southern moiety of Stewart island. 

The processes by which the mountains of New Zealand have been built 
up were distributed over an extremely long period of time. Folding move- 
ments occurred before the Mesozoic aera; the middle and upper Creta- 
ceous in particular, distinguished here as in Europe by the presence of 
a dicotyledonous flora, lies in discordance in many places, but in several 
parts of South island post-Cretaceous folding is also known. The active 
volcanos, the numerous hot springs of North island, and the seismic 
phenomena show that the breaking up of the Cordilleras cannot be regarded 
as yet at an end. 

' Mountains with sharply jagged peaks,* writes Captain Hutton, ' are 
the exception in Switzerland, and the rule in New Zealand. Waterfalls 
are rare in New Zealand ; a few occur up the deep Qords of the south-west 
coast, and some few small ones at the head of the valleys in the great 
ranges. Tet the Alps of New Zealand are quite as bold and steep as those 
of Switzerland; their ravines are even more numerous and deeper. The 
passes are deeper in New Zealand, the valleys much more terraced, and 
the mountains on the whole more extensively covered by loose d^ris than 
in Switzerland. This is certainly truer of Canterbury, Nelson, imd Marl- 
borough, than of Otago. The explanation lies in the fact that the Alps 
of New Zealand are hy far the elder. They have been exposed to the 
action of the weather, at least in part, since the Jurassic period, and many 
of the larger valleys were already excavated, almost to their existing depth, 
before the Oligocene period.' 

The smaller islands which surround New Zealand in the south and 
south-east have a very varied structure. Such observations as exist upon 
them have been brought together by Meinicke and Button^. Since 

pp. 89-48. In this work some remarkable observations are recorded on the passage of 
the stalk of Macrotaeniopteris through seyeral beds of sand ; this suggests that we may 
here have beds of aeolian origin. In the interior of the chain also the syntaxis does not 
take place at an acute angle, for at the northern half of the elbow-like bend of lake 
Wakatipn M^Eiaj observed a strike from north to south (District West and North of Lake 
Wakatipu, tom. cit., p. 118 et seq.), and the south-south-easterly direction occuib first 
on lake Te Anau, which lies on the boundary between ancient schist and gneiss (Coz» 
Report on the Geology of the Te Anau District, op. cit., 1877-1878, p. 110 et seq.). 

^ C. E. Meinicke, Die kleinen Inseln im Sdden und Sfidosten von Neu-Seeland, 
Feterm. Mitth., 1872, XYIII, pp. 222-226 ; F. W. Hutton, On the origin of the F^una 
and Flora of New Zealand, Ann. Mag. Nat. Hist., 1884, 5th ser., XIII, pp. 425-448, and 
1885, 5th ser., XV, pp. 77-107 ; in particular in the latter part, p. 80 et seq. 



CH, HI] AUSTRALIA 149 

nothing in the nature of a general view can be gained from these widely 
separated localities, the following brief account may suflSce. 

From the Snares islands Armstrong obtained basalt, decomposed quartz 
porphyry, and jasper. Auckland island, according to Hector, consists of 
granite, Tertiary sandstone, and volcanic rocks: this accords with the 
specimens collected by Armstrong. On Macquarie island, Scott found 
greenstone and amygdaloids with mesotype and analcime. Oa/mpbeU island, 
according to Hector, presents blue slates and limestone which resemble the 
older Mesozsoic rocks of New Zealand, as well as Cretaceous with flint, and 
Tolcanic rocks ^. The Antipodes islands consist of dolerite and phonolite. 
BourUy island, according to Norman, appears to consist of granite. On 
Chatham island, according to Haast and Travers, mica-schists, Miocene 
limestone, and volcanic rocks occur ^. 

2. Australia. An ancient folded range accompanies the whole of the 
eastern coast of the Australian continent It rises in Tasmania in wild 
highlands covered with great lakes, projects from Bass strait in the form 
of numerous islands, and on the other side, in the south-east of Australia, 
attains its loftiest summit, over 7,000 feet in height ^. It proceeds along 
the east coast, rises in lat. 17^ 30^ S. to over 5,000 feet in the Bellenden 
Ker, and then decreases considerably in height. We shall find traces 
of it, however, up to the extreme north of York peninsula and across 
Torres strait. 

This great folded range may be followed through 34| degrees of latitude, 
but its southern end, at South cape in Tasmania, is defined by a fracture, 
and its northern end is not known. It bears no general name ; following 
CSarke, who has done so. much for our knowledge of Australia, we will speak 
of it as the AvMralian cordiUera. This name, however, ceases to be 
applicable to some of the elements into which it breaks up, though 
these are closely connected among themselves. In the southernmost 
part of the continent the cordillera sends off a hilly range, which curves 
first in an arc and then runs from east to west through Victoria. 
Selwyn and others, however, showed long ago that even in this east 
to west ridge the foldings present the same north and south strike as on 

^ Filhors statements as to the age of the limestone in Campbell island are unfor- 
innately bo contradictory that I mnst refrain from making nse of them; H. Filhol, 
Mission de l*ile Campbell : Constitution g^ologique de llle, Compt. Rend., 1876, LXXXIV, 
pp. 20^205, and Rapports g^ologiques et zoologiques de Tile Campbell avec les terres 
anstrales avoisinantes, op. cit., 1882, XCIV, pp. 563-565. 

* In the Tertiary deposits of Chatham island Hatton recognises the Pare6ia system of 
New Zealand ; Quart. Joum. OeoL Soc, 1885, XI^I, p. 209. 

* K V. Lendenfeld giyes mount Townshend with a height of 7,256 feet as the highest 
pointy whereas its neighbour mount Kosciusko or Mueller's peak (7,176 feet according to 
Neomayr) is usually so regarded ; Lendenfeld, The Glacial Period in Australia, Proo. 
Linn. Soc. N.S.W., Sydney, 1886, X, p. 47. 



150 THE BORDERS OF THE PACIFIC OCEAN [PABTin 

the east coast, so that here the course of the range is perpendicular to 
its structure \ 

Broad plains succeed the cordillera on the west ; then comes, north of 
the Darling river, a mountainous tract, formed by Barrier range and Grey 
range, which also runs north and south. This is again followed by lowland, 
and then we meet the much more important chains situated on the east, in 
part also on the west, of Spencer gulf, lake Torrens, and lake Eyre. To 
the west of these chains is a broad table-land .which extends as far as the 
west coast of the continent. 

The south-western part of this table-land is formed of granite and 
gneiss, with an undulating surface, and along the coast it breaks off in 
a long scarp. This scarp is known in the south as the Darling range, 
Gregory, in his laborious explorations, traced it for more than nine 
degrees of latitude, from cape Beaufort in Flinders bay up to the river 
Gascoyne. In the south it presents a steep face, 800 to 1,200 feet high for 
more than four degrees of latitude, while at a greater distance from the 
sea the table-land reaches a height of 1,400 to 2,000 feet. At the foot of 
the scarp lies a strip of flat land with rare extrusions of basalt, and in front 
of this again, in the extreme south, another gneiss zone running from cape 
Leeuwin to cape Naturaliste. Further to the north, from about lat. SI"" S. 
a narrow gneiss zone again appears in front of the scarp ; it recedes from 
the latter more and more to the north-north-west, including in this manner 
the coal-field on the river Irwin. The same zone of gneiss reappears 
between the river Greenough and Murchison. It is possible that it forms 
Edel land and the regions around Shark bay *. 

The section which Gregory has drawn from west to east in lat. 25° 15' S. 
represents the table-land as consisting of gneiss and granite overlaid by 
metamorphic schists ; while to the west of it lies a Palaeozoic series from 
which Carboniferous fossils have been obtained, and this is covered by 
Mesozoic beds with Ammonites and Trigonias. Since these observations, 
Huddlestone has shown that, according to the collections made by Forrest, 
a great zone of Carboniferous limestone may be recognized on the Kennedy 
range as far as lat. 24"^ S., which extends west of the ancient rocks towards 
the north : the Mesozoic fossils of western Australia have been examined 
by Moore and Neumayr. According to Neumayr's results only a part of 
the middle Jurassic, the zone of Stephanoceras Humphriesianumf can as 
yet be definitely shown to exist here or elsewhere in Australia. The zone, 
however, recalls in a truly astonishing manner the corresponding beds of 

* A E. C. Selwyn, Geology of the Colony of Victoria, in 'The Colony of Victoria in 
Aastralia, its Progress, Resources, &c.,* pablished for the International Exhibition in 
London, 1862, 8vo, Melbourne, 1861, p. 185. 

' F. T. Gr^ozy, On the Geology of a Part of Western Australia ; Quart Joum. Geol. 
Soc, 1861, XVII, pp. 475-488. 



CH.m] SOUTH COAST OP AUSTRALIA 161 

Europe. The general characters of all the species are similar, and some of 
them are identical ; even the nature of the rock, which is a reddish-brown 
oolite, is the same. The beds overlying this zone have afforded a number 
of upper Cretaceous species which have been described by Moore \ 

We may consider the long face of the table-land to which the Darling 
range belongs as a fault scarp^ and the Carboniferous and Mesozoic beds, 
wiiich lie in front of it on the west, as downthrown zones. This is an 
interpretation which in many similar cases elsewhere has alone proved 
capable of explaining all the facts. 

The surface of the table-land, as far as it is known, consists only of 
granite, gneiss, and ancient schists, and is widely covered with sandstone. 
The sandstone contains no fossils, and its beds occupy a very large part of 
the interior. We will call it with Daintree the Desert mndstane. Forrest, 
who penetrated into the interior from the west coast in lat. 29"* S. nearly 
as far as long. 129'' K, also encountered nothing but granite and sandstone. 

The south coast of Australia is particularly characterized by the large 
part taken by marine Tertiary beds in its formation. Clarke has emphasized 
the remarkable fact that while marine Tertiary sediments occur with so great 
an extension on the south coast, yet so far not a trace of such deposits has 
been met with on the east coast in its course from cape Howe to cape York, or 
on that side of the Australian continent which is known in greatest detail ^. 

All these Tertiary beds are horizontal ; their upper limit lies only a few 
hundred feet above the sea ; the fossils they contain show that they are of 
various age. The lower beds contain very few species known to occur 
in the existing seas. The Australian palaeontologists have attempted, in 
q>ite of Duncan's objections, to apply the European terms Eocene, Miocene, 
and Pliocene to the various subdivisions, according to the percentage of 
existing species they contain ^ 

In the Great bight, in the St. Vincent gulf, and in the fluvial region 
of the Murray, these deposits penetrate far into the continent Tate has 
given an instructive account of their mode of occurrence in the Ch*eat 
bight] to this I am indebted for the foUowing account^: — 

^ W. H. Haddleston, Notes on a Collection of Fossils and of Rock-specimens from West 
Ansixalxa, North of the Gkucojne Riyer, Qnart. Jonm. Geol. Soc., 1883, XXXIX, pp. 
582-595, pi. ; C. Moore, On Australian Mesozoic (Geology and Palaeontology, op. cit., 
1870, XXYI, pp. 226-268, pL ; M. Nenmajr, Die geographische Yerbreitung der Jura- 
formation, Denkschr. k. Akad. Wiss. Wien, 1885, L, p. 117 et seq. 

^ Bev. W. B. Clarke, Remarks on the Sedimentary Formations of New South Wales, 
iUostrated by References to other Provinces of Australia, 4th ed., 8yo, Sydney, 1878, 
p. 7 et passim. 

' A snmmaiy is given by J. £. Tenison Woods, Physical Structure and Qeology of 
Aostralia ; Proc. Linn. Soc. N.S.W., 1888, VII, p. 880 et seq. 

* R. Tate, The Natural History of the Country around the Head of the Oreat Australian 
Bight; Trans, and Proc and Rep. Phil. Soc., Adelaide, South Australia, for 1878-1879, 
8to, Adelaide, 1879, pp. 94-128, pi. 



152 THE BORDEES OF THE PACIFIC OCEAN [part in 

The great Tertiary plateau is framed round by granite, gneiss^ and 
ancient schists. These rocks are seen in the west near Culver point (long. 
124'' 4&' E.), in the interior of the country near Boundary Dam (lat. 
29** 20' S., a little to the west of long. 129** E.), then in the east near 
Coldea waters (long. ISl** 50' 81'' K, lat 80' 20' S.), and near Kdinga 
(long. 32" 7' 11" E/, lat. 80' lO' 25" S.). They reach the coast near Fowler's 
bay (between long. 182*' and ISS*' E.). They also form the foundation of 
the whole of Eyre peninsula, which is surrounded only by a narrow border 
of recent marine sediment. 

Within this outer frame stands the plateau which reaches the sea in 
a gentle curve between Culver point and Fowler's bay, thus forming the 
inner limit of the Great bight. 

From Culver point for about a hundred miles to the north-east the 
southern edge of the plateau forms a wall descending vertically into the 
sea ; then the edge recedes a little, a flat known as Boe's plains lies in front 
of it, and the cliff bears the name of Hampton range ; at Wilson's bluff near 
Eucla (long. 129'' E.) it once more reaches the sea, and from here to the 
head of the Great bi^t (long. 131° E.) it again forms a vertical, often 
indeed overhanging wall, which, bordered by no intervening flat land, 
plunges directly into the sea. This is the well-known Bunda diffs. From 
the head of the Great bight up to Fowler's bay, sand dunes and recent 
marine deposits rest against the cliff, and in great part conceal it from 
sight. 

The great tabular Tertiary mass, bounded in this fashion, is named by 
Tate the Bvmda plateau \ on the maps a part of it is known as the 
NvUdbor plain. 

The Bunda cliffs at their western end are 250 feet high, at the eastern 
only 155 feet. They reveal three groups of strata; at the top a bed of 
hard limestone, brown, grey, or reddish in colour, beneath this loose yellow 
Folyozoan limestone, and finally a white chalky rock, with strings of dark 
flint, which presents much resemblance to the white chalk of England. 
This lowest member has the greatest thickness. Near Euda the upper 
limestone is 50 feet thick, the Folyzoan limestone 12 feet, and all the 
remainder is included in the lowest stage. All the members belong to 
the oldest division of the Australian Tertiary. 

The upper limestone forms the entire surface of the Bunda plateau. 
It is an exposed sea floor. No tree, no water, no valley is anywhere to 
be seen. Caves run through the rock ; red clay, residual from solution, 
covers the lower-lying places. It is a desert 'karst' land, and many 
a moving tale is told of the privations and sufferings, sometimes only 
ended by death, of travellers who have ventured upon this vast plain 
without sufficient provision against its dangers. 

The Tertiary deposits which appear on the north side of Kangaroo 



CH. in] AUSTRALIA 168 

island in Aldinga bay, and about a great part of St. Vincent gulf, were 
r^arded by Tate as the continuation of those of the Bunda plateau. 
They rest against older rocks which form a part of Torke peninsula, and 
rise north of cape Jervis in long ranges running north and south. These 
ranges, hitherto unfortunately but little investigated, must now engage our 
attention \ 

The rocks which compose them are all of great age. The few fossils 
hitherto found belong to the lower divisions of the Silurian. In Eyre 
land no definite general strike can be made out. On Yorke peninsula the 
Silurian is known to occur. A long and uniform range commences at 
the northern end of St. Vincent gulf: this is Flvnders range. The 
chain runs with slight curvature along the east side of lake Torrens up to 
its northern end. It attains a height of 3,000 feet Burr, who long ago 
examined the country around mount Arden between Spencer gulf and lake 
Torrens, makes the strange assertion that crystalline rocks such as gneiss 
and mica -schists occur higher in the series than slates, limestone, and 
sandstone ; Selwyn does not express any decided opinion on this point» and 
the relations of the beds cannot at present be regarded as established. 

Mount Norwest, north of lake Torrens, forms the northern end, slightly 
deviating to the north-west, of a branch of Flinders range. Scoular 
encountered here ancient purple slates and projecting ridges of quartadte. 

The mountains, of very inconsiderable height, slope to the east beneath 
a broad plain of clay, containing nodules in which Mesozoic fossils are 
found; this plain surrounds lake Eyre on its southern side, which abounds 
in springa The impervious clay forms the bottom of the numerous salt 
lakes of this region, but it sometimes contains beds of sand in which 
drinkable water occurs. It follows the south- west side of lake Eyre; to 
the west of this great lake slate and quartzite again crop out with 
a steep dip and form Denison range. Here also a well-marked deflexion 
to the north-west occurs. Whether this range is to be regarded as an 
independent structure, or as the continuation of mount Norwest, is difficult 
to determine, owing to the nature of the country \ 

Let us now return to the sea. 

Addaide range is a meridional chain which comes next on the east. 
It begins at cape Jervis, sends off two little branches from its southern end 
to the south-west, and^Jbhen trends northwards nearly up to lake Frome 
(lat 31*" S.); it runs parallel to Flinders range and its summits attain 
approximately the same height Still further to the east lies another 

' The best sommazy is given by Tate, Leading Physical Features of South Australia : 
Anniyersaiy Address of the President to the Philosophical Society of Adelaide for 1878- 
1879, torn, dt., pp. xli-lxzi 

* Gaiin Sooular, Sketch of the Geology of the South and West Parts of the Lake Eyre 
Btnn ; Trans. Proc Roy. Soc. S. Austr., Adelaide, 1887, IX, pp. 89-54, map. 



154 THE BORDERS OF THE PACIFIC OCEAN LPAMin 

meridional chain formed of Barrier and Orey ranges. It presents anci^it 
rocks which appear to be the same as those abready met with. Here we 
are in the midst of a very extensive r^on of marine Cretaceous sediments 
which spread from Queensland to lake Eyre and probably far beyond it. 

I do not propose to discuss the Tertiary deposits of the Murray river in 
detail ; they are more recent than those of the Great bight and are referred 
to the Miocene. Tenison Woods gives 600 feet as the greatest height at 
which they occur. They also extend on the south coast into Bass strait, 
and even reach the southern part of north Gippsland. There also, 
according to Howitt, the greatest height they attain is 600 to 700 feet K 
They do not occur beyond cape Howe. 

We now reach the region of the Great cordillera. It is indeed only 
a continuation, on a larger scale, of the same system of parallel chains 
which we have just passed in review. The mass of the mountains 
is formed of granite and porphyry, crystalline schists, Silurian and 
Devonian rocks, all intensely folded, standing indeed as a rule nearly 
vertical. The strike does not deviate essentially from the meridional 
direction, except in the north, where it becomes north-north-west. The 
Carboniferous lies fairly flat, in any case much less folded than the rocks 
of the high mountains, and all the younger strata may be regarded as 
horizontal. 

In the whole region of the cordillera a very remarkable gap in the 
series of marine sediments occurs above the marine Carboniferous and 
extends as far up as the Cretaceous. In the place of marine formations 
a diversified series of plant-bearing strata appear, among which some 
sandstone beds occur, believed to be of aeolian origin. The variety of the 
floras, already apparent from the earlier investigations of Feistmantel, has 
been brought into the clearest light by Tenison Woods. It is so great that 
we may hope with the progress of these studies to recognize most of the 
members of the Indian Gondw&na series in the region of the Australian 
cordillera *. 

It is a remarkable fact, however, that the flora of the Gondw&na series 
is preceded in Australia by a number of others which have the character 
of the Palaeozoic floras of Europe. The most important which in the 
present state of our knowledge can be distinguished in this region are as 
follows : — 

(a) A Devonian flora with Lepidodendron nothum, which occurs in 

^ A. W. Howitt, Notes on the Physical Geography and Geology of North Gippsland, 
Victoria ; Qnart. Jonm. Geol. Soc, 1879, XXXV, p. 40. . 

' Otto Feistmantel, Die pal&ozoische nnd mesozoische Flora des Ostlichen Australien, 
Palftontographica, SuppL III, Liefl S, 1878-1879 ; and J. E. Tenison Woods, On the fosnl 
Flora of the Cool Deposits of Ausfciulia, Proc. linn. Soc. N.S.W. for 1883, VIII, 1884, 
pp. 87-167, pi. 



} 



cam] FLORAS OF AUSTRALIA 155 

Thoringia in the upper strata of the Cypridina shales, that is to say in 
the upper part of the Devonian of Germany; it is akso known in the 
Devonian of Canada. Carruthers was the first to point out this corre- 
spondenoe^. Species of the genera Cordaites, Sigillaria, Archaeopteris, 
and others, are recorded from the same horizon. 

(b) A flora which corresponds with the European Ouhn. Lqndodendron 
Vdiheirnianv/ni was recognized by Crepin many years ago among Aus- 
tralian fossils. CaUvmitea radiatus, Cycloetigrrui avMrale, Feistmantel 
(which this author admits he can scarcely distinguish himself from the 
Irish Oycloetigma KiUorkerise), Bhxicopteria inaequUatera, and others also 
occur*. 

(e) After the Culm flora the striking correspondence with the floras of 
the northern hemisphere is interrupted for a long period, and strata with 
Oloesapteris Brownia/na rest upon the lower Carboniferous limestone ; they 
afibrd evidence of a glacial period ; in the morainic accumulations of this 
epoch gold has been found in places. These beds correspond with the 
Talchir deposits of India. 

{d) Above the marine beds which overlie these glacial deposits — the last 
marine beds of the oordillera till the Cretaceous is reached — various plant- 
bearing deposits occur which in their lower part still aflbrd Glossopteris ; 
they include the chronological equivalents of the Trias and the Rhaetic 
Whether the equivalents of the Permian are also to be placed here must 
remain for the present uncertain. 

(e) An extensive and important series of strata characterized by the 
genera Thinnfeldia and Taeniopteris includes the Jerusalem beds of 
Tasmania and the Clarence beds of New South Wales; it is also well 
represented on the east coast of Queensland. 

We shall return to this non-marine series later. The deposits 
immediately succeeding are much more recent; they are marine, and 
according to Neumayr's investigations correspond approximately to the 
Aptian stage of the Cretaceous system. 

The inhospitable highlands of Tasmania are a fragment of the 
Australian cordillera. The greater part forms a table-land about 4,000 feet 
high, which is cut up by valleys. To the east it breaks off" sharply along 
the coast Strzelecki laid the foundation of our knowledge of the island ; 
his results have been supplemented by later works ^ The rocks are the 

^ W. Oarrathen, Notes on Fossil Plants from Qneensland, Australia; Quart. Joum. 
GeoL Sec, 1872, XXVIU, pp. 350-354, pi. 

' Teniion Woods, A Fossil Plant-Formation in Central Queensland ; Joum. Roy. Soc. 
N.8.W. for 1882, XVI, 1888, pp. 179-192, pi. 

' P. E. de Strzelecki, Physical Description of New South Wales and Van Biemen's 
land, Sto, London, 1845 ; Tenison Woods, A Physical Description of the Island of 
Tumania, Trana Roy. Soc., Victoria, Melbourne, 1883, XIX, pp. 144-166. On the 
mitenhed between the river Tamar and Port Sorel river the strike diverges considerably 



156 THE BORDEES OF THE PACIFIC OCEAN [PAKrra 

same as those composing the cordillera on the mainland of AustraliA^ but 
eruptive greenstones belonging to the later part of the Mesozoic period 
occur here in greater profusion. Granite, ancient ciystaUine schists, and 
Silurian beds are exposed in steeply upturned zones with a north and 
south strike; the Carboniferous, of marine as well as extra-marine 
development, lies flat upon them ; then follow more recent plant-bearing 
beds; the horizon with Thinnfeldia is also represented, and all these 
deposits are surmounted by sheets of greenstone, which form a large part 
of the surface of the table-land. Marine beds of the Mesozoic period are 
entirely absent; Tertiary marine deposits corresponding to those of the 
Great bight are known on the north coast at a trifling height above 
the sea. Tertiary basalts occur both in the north and south of the island. 

The rocks of the cordillera, locally overlain by Tertiary beds, now form 
the islands of Bass strait, and if we follow their strike to the mainland we 
reach the Australian Alps, the highest part of the cordillera. R von 
Lendenfeld has described their structure in detail. Although in the south- 
east of Australia the principal range of mountains bends through an arc 
into an east and west direction, yet the structure, as we have previously 
observed, does not correspond with this direction. It is evident from 
B. von Lendenfeld's description that the arc breaks up into a number of 
transverse chains running meridionally or converging slightly to the south ; 
these show the true strike of the steeply upturned zones of granite, gneiss, 
and Silurian in this region. In the Bogong mountains there occurs in addi- 
tion an extensive eruptive mass of a basic rock, probably of Devonian age. 

The tectonic trend-lines are slightly convex towards the east and 
indicate folding from the west. This folding is older than the Carboni- 
ferous ^. 

The mountains proceed with the same structure from the east of 
Victoria to New South Wales, running fairly parallel with the coast, which 
descends rapidly to a depth of more than 2,000 fathoms. The geological 

from the meridian (N. 20''-30'' W) ; Nonnan Taylor, Notes on the Geology of the Wert 
Tamar District, Tasmania, Trans, and Proc. Roy. Soc. Victoria, 1880, XVI, p. 156. 
Many obseryera mention faults in Tasmania. Harrison writes on the region between the 
Derwent and niount Wellington as follows : ' As a veiy homely illastration we may 
suppose a set of wooden cubes to be laid out upon a yielding foundation, say a 80& 
cu0hion, so that the surface of the whole represents a perfectly level superficies. Anon, 
and some disturbing force changes the horizontal plane of each cube into a gently sloping 
incline, and forms at every joint a diminutive escarpment. If we can only imagine that 
some molten substance, such as wax, has been forced through the various interstices 
from beneath, so that its overflow partially fills up the miniature valleys, we shall have 
a model representation of Hobart Town with its sandstones, dislocations, and eruptive 
rocks ; ** T. Harrison, Notes on the Geology of Hobart Town, Trans. Proc. Roy. Soc., 
Victoria, VI, 1865, p. 138. 

^ R. V. Lendenfeld, Forschungsreisen in den Australischen Alpen; Peterm. Mitth., 
Erg&nsungsheft, Nr. 87, 1887, 87 pp., maps. 



c3H.ni] EAST COAST OF AUSTRALIA 157 

msp of New South Wales by Clarke and Wilkinson shows that nearly 
meridional zones of granite and steeply folded Silurian, accompanied by 
a number of similar zones of Devonian and ancient porphyry, form thie east 
coast nearly up to Bateman's bay in lat V 86' Aid" S. ; from this point to 
the north, their eastern border, following approximately the meridian of 
long. 150** E., recedes further and further from the coast ; north of Mudjgee 
on the watershed between the Hunter river and the Macquarie in about 
lat 32"* Iff S., the whole cordillera, still directed north and south, disappears 
beneath flat-lying Carboniferous beds ^. 

While the zone is covered in transgression by the Carboniferous, there 
crops out to the east of it, in nearly the same latitude, along the Manning 
liver, on the north border of the same overlying Carboniferous, another 
great zone of granite and- Silurian, also striking north and south, which 
very nearly coincides with the New England range and runs northwards 
to Queensland. The cordillera thus consists here of two folded ranges, 
which alternate with each other; both strike with the meridian. The 
range coming from the south lies between long. 148"* and 150"* £. and 
disappears before it has reached lat. 32'' S. ; the range running to the north 
begins in lat. 32*' S. and its central part lies between long. 151'' and 152"^ E. 
The Carboniferous formation surrounds the end of one and the beginning 
of the other range, and extends along the coast from lat. SS"" 4(y S. to 
beyond Fort Macquarie, i. e. to beyond lat. 31*' 80' S. More recent plant- 
bearing beds overlie it in the south. 

In this intermediate region of horizontal beds Sydney is situated, and 
the most important coal-measures of the colony occur here ; these are the 
Gloesopteris beds. The river-basin of the Hunter belongs to this region. 

Let us again turn our attention to that folded range which begins in 
lat. 82"* S. with the New England range. Daintree's excellent description 
of Queensland well enables us to recognize its peculiar features K In the 
north part of New South Wales this range consists of folded lower 
Palaeozoic beds through the midst of which rise high granite mountains. 
Towards the north the granite seems to disappear; Devonian beds form 
the greater part of the range and an extremely long band of Devonian 
forms the continuation of the mountainous zone, following the course of 
the coast with a gentle curvature up to Shoalwater bay, i. e. to beyond 
lat. 22''3(/ S. According to existing accounts this chain thus extends 

' Greological Sketch-Map of New South Wales, compiled from the Original Map of the 
late Bev. W. B. Clarke by C. S. Wilkinson ; contained in Ann. Rep. Dep. Mines, N.S.W., 
for 1880, 4to, Sydney, 1881. According to Clarke (Remarks, &c., p. 18) the tin-bearing 
gnuute of Queensland and New South Wales is of Devonian age, and consequently much 
younger than the other granite. 

* B. Daintree, Notes on the Geology of Queensland, Quart. Joum. Geol. Soc., 1872, 
XXVIII, pp. 271-817, map ; R. Etheridge, Description of the Palaeozoic and Mesozoic 
Fcsiils of Queensland, tom. cit., pp. 817-859, pi. 



158 THE BORDERS OF THE PACIFIC OCEAN [PAKPm 

from the parallel of 82*' S., through 9| degrees of latitude, to the north. 
Its opposite slopes are formed of different sediments. On the east side 
towards the sea there lies an almost continuous zone of Mesozoic pUmt- 
bearing beds, which begins in lat. 30"* S. in New South Wales and extends 
to beyond lat. 25* S.^ This is the chief region of the Thinnfeldia beds* 
and, as we see, this important area of deposition lies outside that of the 
coal-measures. On the west side of the mountains, however, lies a long 
zone of Carboniferous deposits. It has afforded marine fossils of the 
Carboniferous limestone and plants of the Culm. The beds are thrown 
into very broad undulations. This zone extends from the head waters of 
the Hunter along the west side of the New England range, and Daintree 
bias traced the great outliers running parallel to the coast which form its 
prolongation even as far as between lat. 21*" and 20*^ S. These outliers of 
the extreme north are situated in a granite region to which we shall refer 
presently. 

The Cordillera slopes to the west, beneath the desert sandstone and the 
marine Cretaceous deposits. These form the soil of the deserts of the 
interior up to the gulf of Carpentaria and far away to the west, only 
interrupted here and there by ridges of granite or Palaeozoic sediments. 
At one locality, however, near Maryborough in lat. 25'' SO' S., a little outlier 
of marine Cretaceous appears on the east side of the range resting on the 
Mesozoic plant-bearing beds. 

The chain which begins with the New England range, terminates as we 
have seen at Shoalwater bay: it consists here of folded Devonian. It 
forms the middle part of the Australian coast which projects furthest to 
the east. In its northern part a strike of N. SO** W., corresponding to the 
course of the coast, has been observed in places. On the west side of 
Broad sound a great ridge of granite rocks crops out to the west of this 
Devonian range. Just as the long anticlinals of a folded range replace 
each other alternately, so apparently do these several zones. On the 
southern part of this granite range lie the most northerly patches of 
the Carboniferous zone ; some Devonian and Silurian accompany the range 
in the west towards the plain, but from lat. 22** S. to the north the whole 
coast, as far as it is known, is formed of granite rocks. In this region the 
mountains have lost considerably in height The desert sandstone 
encroaches steadily upon them from the west. 

Battray has described the desert land around cape York. The granite 
mountains extend along the east coast to the north; cape Melville 
(lat. \4t W S.), cape Direction (lat. 12'' 50' S.), then Weymouth cape, Fair 
cape, and others, consist of granite ; the extremity of the peninsula is of 

^ The south part of the Clarence river is described by Stephens, Notes on the Geology 
of the Southern Portion of the Clarence River Basin ; Pzoc. Linn. Soc. N.S.W. for 1888, 
VIII, 1884, pp. 519-531. 



F^ 



CH,in] CAPE YORK 159 

porphyiy. The desert sandstone forms the subsoil far and wide ; on the 
island of Albany it reaches the sea. A red deposit resembling laterite 
rests upon it over large areas ^. 

The islands and reefs of Torres strait consist in no small part of 
granite, and form the continuation of the oordiUera towards New Guinea. 

According to the observations hitherto made, the folded ranges of the 
Australian continent are parts of a mountain system distinguished by 
common characters. They are all more or less meridional, or so arranged 
that with a slight deviation from the north and south direction, such as 
occurs for instance in north Queensland, they form as a whole an arc 
slightly convex to the east. They are older than the Carboniferous and 
in Queensland in particular older than the Culm. The chains rise both to 
east and west, but chiefly to the east of the line of depressions which is 
marked by lake Eyre, lake Torrens, and Spencer gulf. The first is 
Flinders range, the second Adelaide range, the third Barrier and Orey 
ranges, the direction of which is continued far in the north by McSonlay 
range; the last, according to Daintree, also consists of ancient rocks. 
There now follow, to the east, the members of the great cordillera, namely 
a chain which runs from Tasmania through the Australian Alps up to the 
Hunter river in lat. 32'' S. ; a second which replaces the first in lat. 82'' S., 
east of the Hunter, and strikes parallel to the coast up to about lat. 
22'' 3(K S. ; finally a third which follows and replaces the second on the 
west side of its northern end, then decreases in height and, directed to 
the north-north-west, extends beyond cape York through the islands of 
Torres straits to New Guinea. 

The volcanic formations have not yet been mentioned, although they 
play no inconsiderable part in the structure of the cordillera. There is 
here no question of those older eruptive rocks which, within the folded 
Palaeozoic series and in unmistakable association with the auriferous 
deposits, have played such an important part in these countries, but. we 
are concerned with the much more recent lavas of middle Tertiary or even 
later age. These are exclusively basic. On the higher parts of the 
cordillera and its slopes, from Tasmania through Victoria, New South 
Wales, and Queensland, many basaltic streams and sheets are visible, 
resting in places on leaf-bearing beds which cannot be older than the 
middle Tertiary. Daintree has shown that the desert sandstone of Queens- 
land rests upon lavas of this kind. At the same time it must be borne in 
mind that this widely extended aeolian formation is older than the fauna 
of the great marsupials, such as Diprotodon and its contemporaries, which 
are met with in much more recent breccias and in the erosion valleys of 
the desert sandstone. 

^ A Battraj, Notes on the Geology of the Gape York Peninsula, Australia ; Quart. 
Joum. GeoL Soc., 1869, XXY, pp. 297--805. 



160 THE BORDEKS OP THE PACIFIC OCEAN [PABTin 

In Queensland, somewhat north of lat. 21*" S., on the other hand, 
a number of eruptive centres occur which are shown to be of recent age 
by the complete preservation of their ash cones ; in some places their lavas 
have flowed ovet the desert sandstone. Similarly, volcanic formations, 
still bearing their ash cones, are met with in the south, as in that part of 
Victoria where the direction of the mountains lies transverse to the 
structure, or in other words where the elevations but not the folds are 
bent round towards the west. These volcanic formations, bearing all the 
external signs of a recent da£e, are continued through the south of Victoria 
to the west. The cone of mount Gambier is superposed on horizontal 
marine beds of Tertiary age; Woods has described it in detail^. The 
remains not only of great marsupials, but also of the dingo, have been 
found in or beneath the ashes of these volcanos. 

Basic volcanic eruptions have thus taken place in the Cordillera since 
the Miocene period, and in the north, as well as in Victoria, they have 
lasted up to a veiy recent period. 

There is little to say of the deserts of the interior. All that enter- 
prising travellers have seen or collected in these inhospitable regions is 
granite, which appears to crop out at the surface in a weathered condition 
over vast areas; ancient schists; some exposures of marine Mesozoic 
deposits; and a vast covering of desert sandstone. The doubts which 
have been raised as to the presence of marine Jurassic beds in the east 
of the country have already been mentioned ; the same tmcertainty affects 
all the finds hitherto made in the interior; the Cretaceous alone can be 
regarded as definitely identified. This certainly extends from the western 
slopes of the cordillera, or more exactly from the outlier on its east side 
near Maryborough in Queensland, to mount Stuart in Grey range, beyond 
this past lake Eyre, to the north-west nearly to the gulf of Carpentaria, 
and certainly much further still into the interior. 

The explorers who have entered the cotmtry from the north-west coast 
have also encountered only granite and desert sandstone, or possibly basalt 
as well, e. g. on the Victoria river. True Jurassic fossils are not obtained 
tmtil we reach the Glenelg river between lat IS"* and IS* S.; this was 
the locality assigned to the specimens sent to Europe which Neumayr 
described ; they show a striking correspondence with the species collected 
by Gregory and his successors in the more southerly parts of the west 
coast ^. Similar deposits have as yet nowhere been met with in the centre 

^ J. £. Woods, Geological Obaervationt in South Australia, 8to, London, 1862, p. 224 
at seq. 

' Neumayr, op. cit, p. 140 et seq. Depuch island, long. 117'' 44' £., lat 20'' 37' S., 
is said to be a great accumulation of blocks of greenstone rising 514 feet high above 
the flat coral reef of the neighbourhood ; Wickham, Note on Bepuoh Island, Joum. 
Geogr. Soc, 1842, XII, pp. 79-83. 



cH.m] AUSTRALIA AND SOUTH AMEMOA 



161 



or east of Australia, and we most for the present consider these remark- 
able sediments, which recall so strongly the Jurassic deposits of Europe, as 
confined to the west of the table-land . 

Let us now cross to the other side of the Pacific Ocean, within the 
8une parallels of latitude, keeping in mind the results we have obtained in 
Australia and New Zealand. 

On the western border of the Gran Chaeo rise a number of lofty 
narrow chains composed of ancient rocks and running north and south: 
the dip of the beds is very steep. The most easterly of these, near the 
town of Cordoba, as for instance the sierra Ischilin, sierra de Cordoba and 
sierra Cerezuela, oonsist of granite, gneiss, and Archaean schists. Further 
north the sierra de Aoonquija, formed of granite and gneiss, and a large 
number of narrow parallel ranges composed of Cambrian and Silurian 
beds continue this system of meridional sierras as far as Bolivia (I, p. 512). 
FnrUier to the west the series of marine formations is completed by the 
addition of mighty Mesozoic deposits. Recently, Mojsisovics has shown 
that the marine Trias also occurs in the northern part of the Andes, as 
proved by the specimens found by Beiss and Strubel in Peru, and by 
lindig in Chaparal^ 

With the completion of the Mesozoic series we reach the great 
mountaina Beyond them, there lies towards the west the deep valley 
of Chili, then the singular Coast eordiUera, and last the Ocean. 

Now let us cross Australia, but instead of from east to west we will 
take an opposite course and travel from west to east. 

In place of the Gran Chaco or the Pampas, there lie before us the 
deserts of Western Australia ; instead of the lagoons of the western border, 
the sheets of water which extend from Spencer gulf to lake Eyre ; instead 
of the meridional sierras of Cordoba with their ancient rocks, we have 
ninders range, Adelaide range. Barrier and Qrey ranges ; and the place of 
the other parallel chains is taken by the alternating members of the 
Australian oordillera.' For a long distance the further extension of the 
continent is now concealed by the sea, but beyond it, in New Zealand, 
the Meeoz<nc series is completed, the marine Trias is represented together 
with several members of the Jurassic system, and with this completion of 
the series we' reach the great ranges and at the same time a region of much 
more recent folding '. 

Just as the Argentine sierras on the border of the great plain cannot 
be s^iarated from the principal range of the Andes, but constitute with it 



f- 



1^ 



' E.?(m MojaiflOTicn, ArktiflcheTriaBfamien; M^m. AcadLlmp. SoL Saint-P^tenb., 1886, 
7« sfr., XXXni, p. 151. 

' On the greater oompletenen of tiie series in New Zealand, see in pariiciilar Hector, 
tlie (Jeologioal FormatioiiB of New Zealand compared with those of Australia ; Joum. Roj. 
See. N.aW. for 1879, XIII, 1880, pp. 66, 67. 



•OHMMMU 



H 



/ 



I 



t 



162 THE BORDEES OF THE PACIFIC OCEAN [pakp in 

a single xnountam system constructed on a common plan, so all the chains 
from Flinders range to the Australian cordillera, including the longer of 
the two syntactic mountain segments of New Zealand, must be equally 
regarded as parts of a single system similarly constructed on a common 
plan. 

In this connexion we may recaU the remarkable distribution of the 
marine Tertiary beds which border the south coast of Australia, enter into 
the transverse subsidence of Bass strait, and even reach the southern part 
of North Gippsland, but are entirely absent on the whole east coast of the 
mainland, as well as on the east coast of Tasmania, which is bounded 
abruptly by steep cliifs. Clarke consequently has been led to conjecture 
that the continuation of the Australian continent towards the east is cut off 
by a recent subsidence \ 

In confirmation of this hypothesis it may be pointed out that on Lord 
Howe island, between Australia and New Zealand, great bones of terrestrial 
animals have been found which are ascribed to the gigantic lacertilian 
forms Megalania and Notiosaurus. On the Australian mainland these 
genera were contemporaries of Diprotodon and other great marsupials^ 
and thus lived after the deposition of the desert sandstone, or in veiy 
recent times. The existing area of Lord Howe island could not possibly 
have supported such large animals *. 

3. New Caledonia. Many years ago distinguished observers, such as 
Dana and Clarke, had been led to regard New Caledonia as the continua- 
tion of New 2^1and ; the connexion was said to be indicated by the sbnke 
of the north-western peninsula of New Zealand, and by Norfolk island. 
We know now that the direction of the peninsula does not correspond with 
the principal strike of New Zealand. There are certainly, however, some 
features in New Caledonia which reveal a particular resemblance with 
New Zealand. 

Gamier and Heurteau have published detailed reports on the structure 
of this great island : its fossils have been described by Deslongchamps and 
P. Fischer \ The structure is as follows : — 

On the south-west coast, and especially in the southern half of the 

^ Clarke, Remarks on the Sedimentary Formations of New South Wales, illostrated by 
References to other Provinces of Australasia, 4th ecL, 8to, Sydney, 1878, p. 7. 

» R. D. Fitzgerald, Proc. Linn. Soc. N.S.W. for 1884, IX, 1886, p. 1206. 

' Grander, Essai sur la geologic et les ressources min^rales de la NouyelleClalddonie, 
Ann. Mines, Paris, 1867, 6* s^r., XII, p. 1, map; E. Heurteau, Rapport k M. le Ministre 
de la Marine et des Colonies sur la constitution g^ologique et les richesses min^rales de 
la Nottvelle-CalMonie, op. cit, 1876, 7« s^r., IX, pp. 232-454, map ; for the fossilflv 
E. Deslongchamps, Documents sur la g^ologie de la Nouvelle-Cal^donie, BulL Soc^Iinn. 
Norm., 1864, VIII, pp. 832-378 ; P. Fischer, Notes sur les roches fosnlif^res de PArchipel 
Cal^onien, Bull. Soc. g^ol. de Fr., 1867, 2« s^., XXIY, p. 457, and F. Teller in 
E. Mojsisovics, Arktische Triasfiftunen, Mem. Acad* Imp. Sci. Saint-P^tersb., 1886, 7^ ser., 
XXXIII, p. Ill et seq. 






CH. ra] NEW CALEDONIA 168 

island, melaphyre and associated tuflb appear in several places. These are 
followed on the east by an altered rock in which Spi/rigera Wreyi of the 
New Zealand Trias has been f onnd. This is overlaid by Trias shales with 
Paeudomonotis BichTrumdiana and MytUus problemcxticusj that is by the 
equivalent of the middle Trias of New Zealand. These deposits are known 
in the islands of Duoos and Hngon. The Trias zone is followed by a much 
longer narrow zone striking parallel to the coast and formed of a series of 
coal-bearing beds. These beds are known on the coast for a considerable 
distance to the north-west and consist chiefly of sandstone and conglo- 
merate; some of the marine fossils found in them have been identified with 
species of the European lias ; the coal is supposed to be Bhaetic or Idas. 
The zone which follows the coal-bearing beds is by far the most important 
and occupies the greater part of the island; it is an enormous band of 
serpentine and green schists accompanied by dykes of euphotide, which is 
distinguished by its richness in chromite and nickel. This band of 
serpentine forms the tie des Pins and the southernmost part of New 
Caledonia, then the coasts on the north-east side up to Uailu, that is for 
half the length of the island, and on the south-west side up to Mont Dore. 
It then gives place on this side to the sedimentaiy zones already mentioned, 
and reaches the south-west coast further to the north ; finally, striking 
right across the middle of the island, it reaches the most northerly point of 
New Caledonia and is even continued beyond it into the island of Faaba. 
This band of serpentine, striking from the tie des Pins to Paaba, rising in 
many isolated bosses, black, barren, and rocky, is frequently covered by 
dark red day resulting from its disintegration. It often occupies the 
whole breadth of the island and is characteristic of New Caledonia. 

All the rocks hitherto mentioned strike to the north-west in accordance 
with the direction of the island. This renders it all the more strange that 
in the northern part of New Caledonia there are other and older rocks 
which all strike in quite another direction. Heurteau describes them as 
a zone of mica-schists bordered on both sides by slates. The mica-sclusts 
form the fairly large mountain ridge which separates the river Diahot from 
the east coast ; this ridge is directed to the north-west like the band of 
serpraitine and the lengUi of the island itself, but the mica-schists according 
to Heurteau's explicit statement strike across this ridge to the north-east. 
The slates extend to the south nearly as far as Uailu, from which point 
onwards, as we have already mentioned, the coast is formed of serpentine 
down to the southern end of New Caledonia. Crystalline limestone is 
interstiatified with the slate zones on each side. 

Heurteau's description of the valley of the Diahot shows that the strike 
of this ancient schist is actually N. 20"^ K to N. 55^ E. ; it iis solely at the 
north end of the island that the strike of the northern slate zone bends 
more and more into the general strike to the north-west. 

M % 



164 THE BORDERS OP THE PACIFIC OCEAN [partih 

Thus we recognize in New Caledonia two groups of rocks which strike 
in different directions. The first, occupying much the more extensive area^ 
is directed to the north-west like the island itself> and consists of a zone of 
melaphyre and tuff, of IVias, Mesozoic coal-measures, and the great band of 
serpentine. The second group strikes to the north-east, is known only in the 
north, and consists of mica-schists and slate. The first group is a fragment 
of a mountain complex displaying parallel zones ; as to its precise relaticms 
to the second group it is not possible at present to form any opinion. 

The age of the great serpentine band is not known. Such a vast mass 
of serpentine has never yet been encountered at any other point on the 
whole west coast of the Pacific Ocean. Its position would seem to assign 
it to the outer zones of a great mountain range, like the serpentines which 
are known at so many points in the flysch : Heurteau, on account of the 
presence of nickel and cinnabar in both cases, has compared it to the 
serpentine of New Almada in the coast chain of California ^ The more 
significant is the fact that fossils, which he regards as upper Cretaceous, 
have recently been described by Batte from the interior of New Califomia^. 

Recent volcanic rocks have nowhere been observed in the island. 

I do not propose to refer in this place to the coral formations which 
surround New Caledonia, nor to the important calcareous deposits which 
form the parallel series of the Loyalty islands. These will be discussed in 
a later chapter. It is important, however, in view of later descriptions, to 
mention the fact that Tenison Woods has described marine Tertiary fossils 
from Yiti Levu. The species are extinct, but still of tropical character ^ 

There is an extreme dearth of information as regards the islands which 
follow towards the north-west — so far as they do not consist of coral ree& 
or more recent volcanic formations — and only their general direction 
permits us to form any conjecture as to their structural relationa 

Only isolated specimens have been described from New Britavn and 
New Irdwnd. As regards New Ireland, Schleinitz informs us that at 
Carteret and Sulphur harbour the mountains consist of limestone, but 
rolled pebbles show that granite, porphyiy, hornblende, and sandstone may 
occur in the interior of the island. This agrees with liversidge's account, 
who mentions besides porphyiy and diorite, some grey limestone probably 
ancient, from a mountain 2,500 feet high, also somewhat ancient volcanic 
ashes, red jasper, sandstone, epidote, and amygdaloidal lavas. From New 
Britain various volcanic rocks and white limestone are recorded ^ 

^ Heurteau, op. cit., p. 899. 

* Batte, Pzoc. lann. Soc. N.S.W., 1884, IX, p. 681. The genera quoted are BosteUaria, 
Foflus, Fleorotomaxia (?), Belemnites, Nautilus. 

' J. E. Tenison Woods, On some Fossils from Levuka, Viti ; Proc. Linn. Soe, N£.W. , 
1879, IV, pp. 858, 859. 

'« Von SchleinitK, Annal. d. Hjdrographie, 1876, IV, p. 365 ; A. Live^dge, Bocks ftom 
New Britain and New Ireland, Joum. Boy. Soc. N.S.W. for 1882, XVI, pp. 47-51. 



cam] THE SEA OF BAKDA 166 

4. Thi Sea of Bcmday Borneo. As we approach New Owi/nea we 
pereeive ihe importaace of the Tertiary formation in the archipelago of 
Sundn. We have ahready seen that the marine Tertiary sediments on the 
soufli coast of Australia are widely distributed in the Qreat bights and that 
they enter Bass strait, but that the whole east coast of Australia and 
Tasmania up to Torres strait has never yet furnished any trace of 
marine Tertiary beds. In Hall sound. New Guinea, Maday collected 
marine Tertiary shells^ but Tenison Woods maintains that these are entirely 
different from those of the Tertiary deposits of South Australia ^ 
According to Martin's descriptions a large part of the north-western coasts 
of New Guinea is f onned of Tertiary limestone, chiefly distdnguiahed by 
Qrbitoides and lithothamnia, and probably of early Miocene age : it ends 
in difib 200 to 800 meters high. This limestone likewise forms the little 
island of Koor in the south-west^ and several of the adjacent larger islands, 
as well as the little island of 8owek in the north of Qeelvink bay. On the 
island of Lakahia, opposite Merkus-Ort, sandstone with coal-beds is said to 
oeeor^. 

Hie snow-covered mountains in the interior of New Guinea are wholly 
unknown. 

The Malay arc, which runs from Burma through Malacca, the Andamans 
and Nicobars, Sumatra and Java, to beyond Blores (I, p. 468), is difficult to 
follow, since our knowledge of the eastern parts of this long series of 
islands is very slight; it is, however, quite apparent that Sandalwood island 
and Timor lie outside the arc 

Very little is known about Scmdahuood idamd. Ti/mor differs consider* 
ably from the descriptions we possess of the islands lying within the ara 
Since Beyrieh first pointed out the presence of Carboniferous limestone on 
Timor, numerous investigators, Martin and Wichmann in particular, have 
published accounts of the rocks of this island. Ancient schists, tonalite, 
diorite and serpentine are present ; the Carboniferous limestone is widely 
distributed; the occurrence of Trias has not been placed beyond doubt; 
the Tertiary deposits occupy a large area, and appear, as on Java, to rise to 
a height of several thousand feet in the interior of the island \ 

^ 0. S. WilkiiiBon, Notes on a Collection of Geological Specimens collected by 
W. Macleay, Esq., from the Coasts of New Oninea, Cape York, and neighbouring Islands, 
IB Qarke, Bemarks on the Sedimentaiy Formations of New South Wales, 4th ed., pp. 97, 
98; further, J. E. Tenison Woods, On a Tertiaiy Foimation at New Guinea, Froc. 
Unn. Soc. N.S.W., 1877, II, pp. 125-128 and 267, 268 ; further, by the same, Physical 
Structure and Geology of Australia, Proc linn. Soc. N.S.W., 1883, VII, p. 881. 

* K. Martin, Eine Terti&rformabion von Neu-Guinea und benachbarten Inseln, naoh 
Sammlungen von Macklot und von Rosenberg's, Sauualungen des geologischen Beichs- 
Museums in Leiden, edited by E. Martin and A. Wichmann ; Beitrftge Kur Geologie 
Oat-AsienB und Australiens, 1881-1883, 1. Ser., pp. 65-83. 

3 £. Beynch, Ueber eine Kohlenkalk-Fauna auf Timor, Abh. Akad. Berlin, 1864, p. 61 ; 
K. Martin, Die versteinerungafQhrenden Sedimente Timers, nach Sammlungen von 



*• i,i 



166 THE BORDEBS OF THE PACIFIC OCEAN [partih 

Timor does not follow the direction of the great arc and its relations 
with the latter are not known. Let us now turn to the eastern end of 
the arc. 

It is to my mind a very instructive fact that steep folds occur, as 
Verbeek has diown, in the Tertiary beds which lie at the foot of the 
volcanos in the residency of Cheribon in Java K The continuation of the 
micaH9chists of Java has been discovered in Madura, overlain, in this case 
as in that, by the marine Tertiary sediments. The line of volcanos is the 
only indication of the continuation of the arc through Bali, Sumbawa and 
Flores. Tertiary formations, as in Java and Madura, play a prominent part 
in the structure of these islands. The borders of the Banda aea have 
been described by Biedel. From his accounts we gain the most important 
fact that the islands of Romang, Damma, Teun, Nila and Serua form a long^ 
and connected chain of still active volcanos ; as their further continuation 
we must probably regard the volcanic islands of Tuur, the most southerly 
of the Watubela group (south of the Gorong archipelago) ; its last eruption 
took place in the year 1669. These six volcanic islands lie exactly on the 
prolongation of the great zone of volcanos which curves down from Java^ 
and they carry the arc into the immediate neighbourhood of New Guinea. 
The fact is all the more remarkable since outside the arc volcanic 
phenomena are only known at one point, that is on the east side of the 
island of Moa. 

A series of middle Tertiary beds, consisting of sandstone, marl, or lime- 
stone, forms a great part of the islands lying in front of the volcanic chain. 
Leti and Moa are formed of this series, while the next island, Lakor, is 
a recent coral formation. The Luang-Sermatta group is also middle 
Tertiary, as well as the Babar group ; from the latter, however, Mesozoic 
limestone said to be Jurassic is also mentioned. The Tanembar or 
Timorlao group consists almost exclusively of low-lying coral limestone, 
but in the south-eastern part of Jamdena, the largest island of the group, 
late Tertiary land is visible. In the Eei group the great island Nuhujuut 
is middle Tertiary; it rises towards the north to a height of 400-500 
meters. The principal group of Aaru is a recent limestone plateau 
traversed by «alt- water channels, 25 to 800 meters in breadth, which cross 
the island obliquely from coast to coast ; in the south-east recent Tertiary 
land rises to a height of 50 meters \ It is clear from Martin's accounts 
that these Tertiary deposits are a continuation of those of New Guinea. 

Reinwardt, Macklot und Schneider, Samml. geol. Reichsmus. in Leiden, 1881-1883, I, 
pp. 1-^, plate ; A. Wichmann, Gesteine von Timor, op. cit., 1882, II, pp. 1-72, plate. 

^ R. D. M. Verbeek, Over de dikie der tertiaire afzettingen op Java; Veih. Akad. 
Amsterdam, 1888, XXIII, D, 11 pp., plate. 

' J. G. F. Riedel, De sliuk- en kroesharige raasen tusschen Selebes en Papuan 8to, 
'b Grayenhage, 1886, maps. A map of Aaru in Verb. Gesellflcb. f. Erdk. Berlin, 1885, 
XII, pL i. 



CH, m] BORNEO 167 

The two islands lying north-east of Timor, Eetar (or Weetar) and 
K^sar (1,200 meters high), each consist of ancient rocks with others which 
are apparently of Mesozoic age. In like manner, ancient rocks, said to be 
aasociated in places with Palaeozoic or Meeossoic limestone, form the high 
chain which extends from Burn (2,500 meters) past Ceram (also rising to 
2^500 meters) to the Seranglas and Gorong archipelago. Middle Tertiary 
beds accompany this range, and form the island of Amboina. As on the 
Iiawadi, the Nicobars, Smnatra, and Java, so now on the east coast of 
Ceram also, we meet with springs of petroleum ^. 

The middle Tertiary series, distingmshed in some of its members by 
coal beds, or by Nummolites, Orbitoides, peculiar lithothamnia and the 
genus Cydodypeus, is assigned by Verbeek to the Eocene in part, but is 
regarded by Martin as Miocene. Martin has found Tertiary limestone ia 
Amboina, Buton, and Madura, and notably in Borneo, where it is widely 
distribnted ^. 

It cannot be denied that the volcanic chain which extends from Bomang 
to Tuur presents a close resemblance to the volcanic arc on the inner side 
of the Lesser Antilles. The Tertiary islands and the recent limestone 
plateaux similarly correspond to the succeeding outer arc. The Banda sea 
would thus appear to be homologous with the Caribbean ; and the Arafura 
sea^ lying on the foreland, with the gulf of Mexico or the adjacent part of 
iJie Atlantic Ocean. 

Although Mesozoic or Palaeozoic sediments are frequently mentioned 
as occurring in this region, yet as far as my information extends no organic 
remains, older than the Tertiary period, have hitherto been met with except 
in. Borneo, 

We possess a geological map of this great island, constructed by 
Schtoaner and Gaffron in the years 1843-1848 ; it comprises the southern 

^ Biedel, op. cit., p. 86. On the petroleum of the east coast of Sumatra see Everw^n, 
Jaarb. M^nw. Ned. O. In<L, Y, p. 186 ; in North Sumatra, op. cit., IV, a, pp. 15-38 
and 188 ; in Soerabaja, op. cit., IV, h, p. 118 et passim. 

' K. Martin, Die wichtigsten Daten unserer geologischen Eenntniss vom nieder- 
l&ndiflch-ostuidischen Archipel, B^dr. tot de Taal-, Land- en Volkenkunde van Ned. 
Indie ; uitgeg. vanw. het kon. Inst, &c., ter Gelegenh. van het VI. intemat. Congress d. 
Orientalist, te Leiden, 8vo, 's Gravenhage, 1883, pp. 17-34; cf. also D. Schneider, 
(SeologiBche TJeberdcht flber den holl&ndisch-ostindischen Archipel, Jahrb. k. k. geol* 
Reichsanst., 1876, XXYI, pp. 113-134, maps. For Tertiary formations see in particular 
K. Martin, Nene Fundpunkte von Tertiftr-Gesteinen im indischen Archipel, nach 
Sammlungen von Homer, Eorthals, Macklot, MtQler und Reinwaidt, Samml. geol. 
Beichamns. in Leiden, I, pp. 131-179 ; also A. B5hm, Ueber einige tertiftre Fossilien von 
der Insel Madura, Denkschr.. k. k. Akad. Wiss. Wien, 1882, XLV, p. 359 ; on the disputed 
question as to the age of these deposits I must refer the reader to Verbeek, Boettger, 
Geyler, and C. von Fritsch, Die Eocftnformation von Borneo und ihre Versteinerungen, 
FiUaeontographica, 1875, Suppl. Ill, Heft 1, and Yerbeek, Boettger, and E. von Fritsch, 
Die Terti&rfonnation von Sumatra und ihre Thierreste, op. cit., 1880, Suppl. Ill, 
Heft 8-11. 



168 THE BOEDERS OP THE PACinC OCEAN [part in 

area and extends a little beyond the equator; there are also a series of 
detailed descriptions by Dutch mining engineers and an instructive general 
account of the whole by E. Martin \ 

The first explorers, Homer and Schwaner, had already recognized that 
in the interior of Borneo mountains exist formed of granite, serpentine, 
crystalline schist and other ancient rocks, which strike in a similar direc- 
tion to the hill-ranges of Celebes and Halmahera; the hollows between 
them are, however, filled with Tertiary sediments. Martin shows how 
Pechel and Wallace had independently of each other perceived this curious 
homology of the three adjacent islands which yet finds so little expression 
in the outlines of Borneo '• 

Besides this Tertiary transgression, Verbeek has recently made known 
in Borneo fossilif erous marine deposits of the Cretaceous period, oocurriji^ 
on the Seterocang, a tributary of the river Eapoea which fiows down to 
the west coast. Amongst the specimens collected by van Schelle, Qeinite 
has recognized some fossils so closely related to Vola quadriooftata^ 
Trigoma limhata^ Ooniomya dedgruita, Hemicuier sublcuyumoaua and 
Hemiaster pUbemSt as well as other upper Cretaceous species, that he 
correlated the deposit with the upper Senonian \ 

Tenison Woods asserts that the coalfields of the west are contem- 
poraneous, in part at least, with the Newcastle beds of Australia; with 
regard to Labuan this is merely a conjecture ; the coal district of Sarawak 
has afforded PhyUotheca cmskralis and Vertebraria sp., that is to say species 
characteristic of these Australian deposits ^. 

Finally, marine fossils, either Devonian or Carboniferous, are also 
present; van Schelle mentions traces of them in the western residency 
and Tenison Woods quotes Fenestella and Stenopora from a limestone in 
the norths 

5. Cochin-ckinai Tonghmg, In recent times our knowledge of the 

^ E. Martin, von Gaffron^s geologische Earte von Stld-Bomeo, Samml. geol. Reichsmw. 
in Leiden, I, pp. 179-193, map ; also op. cit., p. 182 et seq. 

* T. Posewitz* account of the case is perhaps too uncompromiBing, Unsere geologisehen 
Eenntnisse von Borneo ; Jahrb. k. ung. geol. Anst., 1882, YI, pp. 135-162, map. 

^ R. D. M. Yerbeek, Over het voorkomen van gesteenten der krijtfonnatie in de 
residentie Westeiafdeeling van Borneo ; Yersl. en Meded. E. Akad. Wet. Amstextlam, 
Afd. Natuurk., 1884, XIX, 2. reeks, med. pp. 89-43. Martin regards these depodts as 
a tropical modification of a Tertiary formation, but no member so far known among tlie 
Tertiary deposits of Java and Sumatra appears to be so rich in species usually supposed 
to be Cretaceous types. 

« Tenison Woods, The Borneo Coalfields, Nature, April 28, 1885, pp. 588, 584 ; ef. also 
J. Motley, On the Geology of Labuan, Quart. Joum. GeoL Soc, 1858, IX, pp. 54-^7. 

' Schelle, Bericht in Peterm. Mitth., 1885, p. 820 ; Tenison Woods, The Geology of 
Malaysia, South China, &c.. Nature, Jan. 7, 1886, p. 232. Hart Everett mentioBs 
orinoidal limestone from north-east and north-west Borneo, but without any closer 
indication of age ; Report on the Exploration of the Caves of Borneo, Proc. Roy. Soc, 
1880, XXX, pp. 810-821. 



"1 



CH. m] COCHIN^mNA 169 

mainland of Further India has been largely extended by the labours of 
French geologists. The bold expeditions of Gamier have been followed by 
the investigations of Batte and Petiton in lower Cochin-china and Cambodia, 
of K Fnchs on the coast of Annam and in Tongking, and of Jourdy on the 
east of Tongking \ While many questions must still remain open, sufficient 
is ascertained to furnish the following account. 

As in Borneo, recent eruptive rocks are very feebly represented; 
isolated occurrences of such rocks surround indeed the south and east 
eoasts of Further India, but on the. mainland itself they are known only at 
a single locality, quite close to the south coast. A boss of basalt occurs 
at this place, which is situated south of Bien-Hoa. The Tigre, a little 
island on the east coast (N. of lat. 70'' N.), consists of trachyte ; trachyte 
is abo known from Pulo Condore and Pulo Wai in the east and west of 
the southern extremity of Cochin-china. 

A part of Pulo Condore, however, consists of granulite, and this is the 
forerunner of numerous protrusions of this rock which rise near cape 
SainlrJacques and in the neighbourhood of Baria. They are accompanied 
by diorite and are associated towards the north with considerable masses 
of ancient schistose rocka These 'isolated exposures of ancient roeks, 
extending towards cape Saint-Jacques, must be regarded together with 
Pulo C(Hidore as the southern extremity of the long range of hills which, 
starting from the granitic table-land of Laos, descends along the east coast 
of Annam. 

We will first describe somewhat in detail the delta of the Mekong. 

The granulite hills of Baria are continued to the north-west to Bien- 
HoA. Still further to the north-west there rises from the alluvial land 
a ridge of friable sandstone, the first outlier of a denuded sheet of sand- 
stone, numerous remains of which again appear in the north superposed 
upon Rhaetic coal beds ; to the west of the sandstone ridge rises another 
gianulitic mountain near Tay-Ninh. On the west side of the great river 
between Chaudok and Bach Oia a group of granulitic and granitic hiUs 
again occurs, accompanied by the same sandstone; to the north-west of 
this group, western Cambodia as far as it is known is covered over wide 
areas by the friable sandstone, resting upon upturned beds of limestone, 
which are assigned to the lower Carboniferous. The great island of 



* F. Ratte, Note but rindoXJhine, Bull. Soc. g^oL de Fr., 1875-1876, 8* e^r., IV, pp.50d- 
522 ; Petiton, EsquiBse g^ologique de la Cochinchine franfaise, du Cambodge (province 
de Pounat) et de Siam (province de Battambang), op. cit., 1882-1883, 8* s^r., XI, pp. 
384-399, map ; E. Fucha et £. Saladin, M^moire sur Tezploration des gttes de com- 
bvutiblea et de quelqnes-unes dea gites m^tallif^res de Tlndo-Chine, Ann. Mines, 1882, 
8* 8^., II, pp. 185-298, map ; E. Jonrdy, Snr la gdologie de Test du Tonkin, Compt. 
Bend., 1886, CII, pp. 987-939, and Bull. Soc. g^ol. de Fr., 1885-1886, 3« s^r., XIV, pp. 
14-20> plate ; Note compl^entaire, torn, cit., pp. 448-453. 



170 THE BORDERS OF THE PACIFIC OCEAN [PABrm 

Phonkok consiBtB of the same sandstone ; which also forms the Elephant 
mountain situated to the north of the island. 

These heights, and especially the granulitic ridges and mountains 
between Baria and Bien-Ho&, those of Tay-Ninh, and on the other side of 
the Mekong those of Chaudok, form an incomplete boundary about the 
vast alluvial region ; or perhaps they should rather be regarded as repre- 
senting islands and peninsulas united by the alluvial land which now alone 
forms the whole region up to its southernmost extremity. The part played 
by the great interior lakes of Cambodia in this extensive advance of the 
growing land is very peculiar. Fuchs has given a lucid description of 
the process. The sea once covered the area now occupied by the alluvium 
of the Mekong. The mouth of the river then lay far to the east of the 
great lakes near the existing province of Pnom Bach^; the heights of the 
existing province of Eompong Soai, composed of granitic rocks and quarts 
porphyry, separated the mouth from the great lakes which then formed 
the most northerly part of a gulf penetrating deep into the land. The 
advance of the river deposits cut off this part of the gulf from the open 
sea. The river Tonle Sap, which at present joins the Mekong near Pnom 
Penh and unites that river with the lakes, has so trifling a fall that during 
one part of the year it bears the high waters of the Mekong northwards 
into the lakes and for the other part of the year it carries the overflow of 
the lakes southwards into the Mekong. The direction of the waters is 
reversed as soon as the Mekong rises near Pnom Penh to a height of 
7-8 meters above mean level ; but it rises every year to a height of 12-14 
meters above this level. 

There is a great dearth of information concerning Annam. We know 
that a hilly ridge of granite accompanies the coast, that ancient schists 
occur along its eastern side on the river Tourane near Hue, and that 
ooal-bearing deposits are also present which probably correspond with those 
of Tongking. 

The plain of Tongking is much better known. Friable shales with few 
fossils are assigned to the Devonian, but this determination needs confirma- 
tion. A very prominent element is a hard limestone resembling marble, 
which projects in bluff rocks from many parts of the delta and forms 
outside it innumerable reefs and islands. Fossils of the Carboniferous 
limestone have been found in it by Jourdy. Above it lie calcareous shales 
intercalated with sandstone and containing fossils which, according to 
Douville, are closely related to Myophoria Oold/usd of the Keuper. I shall 
shortly have to refer to similar Myophorias obtained by IxSczi from 
Tshung-tien in Yunnan. There now follow sandstone and shales which 
form the coal-bearing series of Tongking. Zeiller has examined the flora 
of these beds, and has arrived at the remarkable result that 10 to 12 species 
correspond with those of the Bhaetic stage of Europe, while an equal 



GH. m] 



THE PHILIPPINES 



171 



number are known in India, on the horizon of either the upper beds of the 
lower QondwAna or the base of the upper Gondw&na. The number of new 
species is comparatively small ^. 

This series of strata is thrown into folds which strike south-west to 
north-east ; they reach the sea, and Jourdy indicates their continuation on 
Hainan, but on what grounds I have no information. According to the 
same observer this system of folds striking north-east is traversed by frac- 
tures which run from north-west* to south-east, at right angles to the folds. 
Deposits of more recent age than the Bhaetic coal-measures are so far 
not known in these regions. 

6. The Philippmes. The observations made in Celebes and Halmahera 
so far known to me are not sufficient to hazard even a conjecture as to the 
strocture of these islands. We only know that ancient rocks such as 
granite and ancient schists are present, and a large number of recent 
volcanos, many of them still active. 

The accounts of the PhiUppvnes are much more complete, but here 
again we are unable to form any definite opinion as to the structural plan. 
Vie Philippines have so far furnished no fossil remains of the Palaeozoic 
or the Mesozoic period. Superposed upon gneiss, talc-schists, serpentine 
and widely distributed gabbro and diabase, lies a sedimentary series which 
consists of a zone of limestone conformably overlaid by coal-bearing beds 
of unknown age, further of Nummulitic limestone, a more recent coral 
limestone reaching a height of 4,000 feet, which I correlate with the 
middle Tertiary coral limestone of the Banda islands and Borneo, and 
finally recent marine sediments. These formations are associated with 
a varied series of volcanic rocks which date at least from the Nummulitic 
period. 

All observers are agreed that the ancient formations are to be regarded 
as forming, not a table-land, but a number of chains. Of these chains, 
however, only fragments are to be seen ; the sea conceals a large part, and 
another part is smoth^^ up under ashes and tuff ejected from the recent 
volcanos. The outlines of the islands do not everywhere correspond with 
the strike of the ancient formations, but have been affected in many ways 
by the presence of volcanic masses; this is particularly the case in the 
southern part of Luzon. 

In 1878 B. von Drasche, in a compendious description of Luzon, 
represented the Philippines as a number of chains which in the north of 
this island are closely crowded together and run nearly north and south, 
but diverge in virgation towards the south and south-west. ' In north and 
central Luzon, the north and south direction predominates; in the south 

' B. 2Seiller, Examen de la Flore foasile des couches de charbon du Tong-king, Ann. 
Mines, 1882, 8« ser., II, pp. 292-852, plate ; also Bull. Soc. g^l. de Fr., 1882-1883, 3« s^r., 
XI, pp. 456-461, and 1885-1886, 8« s^r., XIV, pp. 454-463, plate. 



172 THE BOEDERS OF THE PACIFIC OCEAN [part ra 

the islands diverge from one another radially like a fan, and a gradual 
taming round may be observed from north-west and south-east to north- 
east and south- west, so that the directions of the strike in Palawan aiid 
south Luzon, the opposite ends of the fan, are perpendicular to one 
another.' The eastern branches are thereby bent more strongly to the 
east, and Drasche emphasizes their tendency to adapt themselves to the 
course of the coast of Annam in the same way as New Zealand and New 
Caledonia do to the east coast of Australia^. 

Subsequent observations seem to confirm this conception. To arrive at 
a clear idea of the position of the several chains, we must bear in mind 
two conspicuous lines of depression, which as early as 1869 were definitely 
marked by Semper on his general map of the islands^. The first of these 
lines corresponds with the longitudinal depression which runs through the 
west of Luzon from the bay of lingayen to that of Manila, and marks off 
a mountain chain, the sierra de Zambales, from the rest of the island* The 
second line corresponds with the depression which occurs in the eastern 
part of Mindanao, and runs, interrupted by a watershed of no great height, 
from the bay of Butuan in the north to that of Davao in the south* This 
also bounds an independent mountain chain which strikes ahnost north 
and south from cape Surigao to cape St. Augustine. 

The sierra de Zambales in west Luzon was crossed by Drasche in two 
places ; it consists chiefly of gabbro and talc-schist with serpentine ; gteat 
masses of trachytic tuff rest on its eastern slope. Its southern extremity, 
projecting into the gulf of Manila, is however of recent volcanic origin, 
and must be disregarded in considering the direction of the chain. The 
longitudinal depression is broad; it presents recent marine deposits* as 
observed by Cen'teno ; Arayat, an isolated volcano, rises out of its plain \ 
It appears to have been in this more independent branch of the sierra de 
Zambales, which curves in an arc towards Palawan, through Luban and 
the Calamianes, that Tenison Woods observed ancient limestone ^ The 
island of Palawan is formed, according to Centeno, by a long continuous 
sierra. This arc of Zambales-Palawan, the most westerly of the vii^rar 
tion, does, in fact, take a course which corresponds in a striking manner to 
that of the coast of Annam. 

^ R. von Drasche, Fragmente zu einer Greologie der Insel Lason (Pbilippinen), ito, 
Wien, 1878, p. 8. The oldest geological account of the Philippines known to me ib that 
by the mining engineer, I. Sainz de Baranda, Islas Filipinas, 8to, Manila, 1840. In 
1873 Both attempted to give a brief account of them from Jagor's collections ; this 
excellent description, however, has little reference to the tectonic relations; J. Both, 
Ueber die geologische Beschaffenheit der Fhilippinen, in F. Jagor, Beisen in dan 
Fhilippinen, 8vo, Berlin, 1873, pp. 338-354. 

' C. Semper, Die Fhilippinen und ihre Bewohner, sechs Skizzen, 8vo, Wfirzbnrg, 1869. 

' J. Centeno, Memoria geol6gico-mineral. de las Islas Filipinas ; Bol. Com. Mapa geoL 
EspaSIa, 1876, III, pp. 181-284, map, in particular p. 184. 

* Tenison Woods, Nature, Jan. 7, 1886, p. 282. 



CH. m] LUZON 173 

In the north of Lozson two great oordilleras, separated by the Rio 
Grande de Cagayan, extend as far as the latitude of the golf of lingayen, 
thai nnite, and for a certain distance follow the east coast. Further to 
the south the mountainB are very much broken up and overlaid by exten- 
sive volcanic masses, but the strike to the south-south-east appears clearly 
in a long band of Nummulitic limestone interbedded with trachjrtic tujBT, 
which may be traced from the province of Bulacan across the -Laguna to 
Ifajaijay. Bichthof en was the first to note its existence \ 

Drasche supposed that the outline of Masbate was due to bifurcation, 
and this would correspond with the structure of Porto Rico in the Antilles ; 
but the structure of Masbate is unknown. 

Zebu has been described in detail by Abella. The oldest rocks are 
dkxrite and dioritic tuff; they form two considerable masses in the centre 
of the island and a smaller mass at its southern end, and are surrounded by 
beds containing Nummulites. More recent limestone forms the remainder 
of the island, that is, by far the greater part; in some places it lies 
horizontal, in others it is upturned like the Nummulitic beds. The lignites 
of Zebu have also been placed in this group, which would thus include 
deposits of various age. The strike of the rocks corresponds on the whole 
with the outline of the island, and thus to the direction of one of the 
middle branches of the virgation '. 

The coal beds also make their appearance, according to Centeno's 
observations, on the adjacent island, Isla de Negros, and traces of them 
are known on the western part of Mindanao, north of the bay of Sibaguey. 
It is possible that a part of western Mindanao may prove to be the con- 
tinuation of Zebu and the Isla de Negros. 

The gold-bearing hill range of Pigholugan, near the gulf of Macajalar 
in Mindanao, that is nearly in the middle of the north coast of the^ island, 
consists, according to Abella, of ancient quartzite and clay slates which 
strike north-north-east and south-south-west ^ 

The eastern sierra of Mindanao, which lies beyond the two bays of 

Butoan and Davao (or Tagloc), was visited by Montano ; from its northern 

extremity it proceeds to the south-south-east, and then turns due south ; 

it consists of greywackes, ancient schists, primitive limestone, serpentine, 

quartz, porphyry, and melaphyre. This mountain range must probably 

be regarded as the continuation of the island of Leyte ^. 

^ F. von Richthofen, Ueber das Yorkommen von NtmmiTiliteiiformen auf Japan nnd 
den Fhilippinen ; Zeitscbr. dentsch. geol. Ges., 1862, XIY, p. 358 et seq. 

* E. Abella y Ctouiego, Bapida descripcion fisica, geol6gica y mineral de la Isla de 
Cebu ; Bol. Com. Mapa geol. EspaSa, 1886, XIII, pp. 1-187, map. 

* Id., Memoriaaceroa los criaderos auriferos del segondo distrito del departamento de 
Mindanao ; op. cit., 1879, VI, pp. 83-79, maps, in particular p. 60. 

* J. Montano, Rapport k M. le Ministre de I'lnstruction publique ear nne minion aux 
Hes Philippines et en Malaisie ; Archives des Missions, 1885, 3® s6r., XI, pp. 271-479, 
maps, in pariicalar pp. 271-277. 



*1 



174 THE BORDEES OF THE PACIFIC OCEAN [PABim 

Still farther towards the exterior lies a second band of coal-bearing 
beds which strikes to the south-east, and has been traced by Genteno from 
the region of Caramuan, across the island of Batan, and across Sngad in 
south Luzon into the island of Samar. 

If we transfer these observations to a map, we attain a virgation from 
south-west in Palawan to south-east in the zone of the coal beds of south 
Luzon and Samar. 

It is not improbable that the most westerly arc, Zambales-Palawan, is 
continued to Borneo. Such a continuation is less certain for the mountains 
of west Mindanao and Basilan ; Yolo consists, according to Itier, of coral 
formations resting on basalt^. The prolongation of the eastern chain of 
Mindanao towards the south is only indicated by volcanos : on the west 
coast of the bay of Davao, the cone of Apo rises, according to Montano's 
measurement, to a height of 3,143 meters ; it forms the highest peak of the 
Philippines. The volcanic zone would now seem to extend through Butulan 
and the island of Sangir to the volcanos of northern Celebes. 

It was generally assumed by earlier writers, that a single zone of 
volcanos extended in an arc through the Philippines to the north ; but the 
case is not so simple. Centeno considers that there are two chains of 
volcanos which should unite in south Mindanao, The western chain 
begins on the Buguias, north of lingayen bay, and includes the volcanos 
of Arayat and Taal, Canlaon on the Isla de Negros, the island of Faego, 
the volcano de Macaturin in Mindanao, and the volcano of Cotabato; this 
line has been named the system of the Taal. The eastern chain is the 
system of the Majon (Albay), and extends from the Isaro in south Luzon 
past Albay and Bulusan, the solfataras of Leyte, the volcano Camigoin 
north of Mindanao, which first arose in 1871, then to Apo and Butulan. 
The solfataras in the trachytie mountains of the island of Bitisan, which 
have recently been described by Abella, might then be regarded as inserted 
into this zone. Abella^ however, conceives the continuation of this line as 
drawn not through Camiguin, but further east, following the direction of 
the island of Leyte towards Mindanao ^ and Drasche's geological map 
of south Luzon shows so many volcanic eruptions of such various descrip* 
tion, that it is difficult to find definite guiding lines in this region. The 
clearest line of this kind makes its appearance in south-east Luzon between 
the Isaro and the Bulusan, and is continued towards Biliran ; it lies com- 
pletely in the strike of this part of the virgation \ 

^ J. Itier, Extrait d'une description de Tarcliipel des lies Solo ; Bull. Soc. g^gr., 
Paris, 1846, 3« s^r., V, pp. 311-819. 

' From Moiitano*8 last description one would certainly be inclined to doubt that lake 
Mainit on the peninsula of Surigao (N. Mindanao) is really a crater. 

* Drasche, p. 72, assigns to this line the closely adjacent volcanos of lAbo, IsanS, Iriga^ 
Masaraga, Albay, and Bulusan. 



cam] FORMOSA 175 

Farther to the north of Luzon, in about lat. 18* !(/ N., where Cabo del 
Engaiio begins to jut out, stands the volcano de Cagua. Towards the 
north-west there follow the volcanic island of Babuyan, then the Batanes 
islands, and finally ihe great island of Formosa. 

7. FormoBa arid the Lm-Kiu ialands, Formosa is still unfortunately 
bat little known; the observations of Swinhoe, Bichthofen, and Tyzack 
are confined to isolated localities in the north ^. Monkey mountain or 
Taku Shan, on the west coast, has been described by Ouppy ' ; Eleinwachter 
has visited the south part of Formosa ^ An important mountain range 
which, according to some statements, reaches a height of .10,000 feet, 
according to others of 12,000 feet, extends through the eastern part of the 
island towards the north and north-north-east. The east coast is in places 
very steep, and near Chokeday (lat 24* 10^ N.), according to Bichthofen, 
it descends precipitously from heights of 6,000-7,000 feet into the deep sea. 
On the other hand, the whole western side of the island is low, and the 
plain is only interrupted by isolated hills formed of coral limestone. 

Kleinwachter reached the slopes of the Eueili Shan, the highest peak 
(about 9,000 feet) in the southern half of the mountain chain, and he found 
that this consists of folded hornblende-schists and clay slates, together 
with quartz porphyry. The pebbles of the river Tamsui show that further 
north granite also enters into the structure of the range. Against these 
mountains there rest in the south, first quartzose sandstone, and then coral 
limestone, of which the two southern promontories of Formosa consist. 
We have already mentioned that coral limestone forms the isolated ihoim* 
tains of the western plain ; it is assigned, as a rule, to the Tertiary ; and it 
also forms the island of Lambay, lying off the west coast. In the isolated 
Monkey mountain, cavernous limestone with Scutella, Balanus, and corals 
rises 1,100 feet above the sea in beds which dip SS"" to 40*" E. In the north 
of the island, on the Eolung river, middle Tertiary coal-seams are being 
worked; they lie in a synclinal with steeply inclined limbs; these, like 
those of Monkey moimtain, show that Formosa must be included among 
iJie regions which have experienced folding in recent times. 

^ Bobert Swinhoe, Notes on the Island of Fonnosa, Journ. Boy. Geogr. Soc, 18B4, 
XXXIV, pp. 6-18, map ; F. von Richthofen, Ueber den Gebiigsbau an der NordkOste von 
Formosa, Zeitschr. dentsch. geol. Ges., 1860, XII, pp. 532-545 ; D. Tyzack, Notes on the 
CoalfieldB and Coal-mining operations in N. Formosa, Trans. N. EngL Inst. Eng., 
Newcastle, 1884-1885, XXXIV, pp. 67-79, map ; and J. A. Leboor, Notes on some fossils 
from N. Formosa collected by Mr. D. Tyzack, tom. cit., pp. 81, 82. Arthur Comer 
mentions a Palaeozoic fossil, Monotis Hatcni, from the neighbourhood of Monkey mountain, 
but this discovery is so far unconfirmed ; A Journey in the Interior of Formosay Proc. 
Geogs. Soc., 1874-1875, XIX, p. 515. 

' H. B. Guppy, Some Notes on the Geology of Takow, Formosa ; Journ. N. China Br. 
& As. Soc, Shanghai, 1881, new sen, XVI, pp. 13-17. 

3 G. Kleinwachter, Researches into the Geology of Formosa ; op. cit., 1884, new ser., 
XVin, pp. 87-53, map. 



176 THE BOEDERS OP THE PACIFIC OCEAN [paotih 

At the mouth of the river Tamsui, Bichthofen observed trachyte and 
horizontally bedded trachytic tuff. The great solfataras, north-west of 
Eolung, probably also proceed from trachyte K These trachytes are the only 
recent volcanic rocks known in Formosa, and the early statements as to 
the presence of active volcanos in the island have not so far been confirmed. 
Eleinwachter, it is true, mentions that on the flat coast near Langchiao 
flames often break forth from the baked ground, but he conjectures that 
they are due to gases which accompany petroleum. The Chinese statement 
that in the year 1722 flames and mud burst forth on the Chih Shan or 
Pineapple hill, near Takow, may perhaps be explained in the same way. 
Petroleum is, in fact, known to occur in the north near Tangshui below 
Tamsui. All these localities belong to the west coast. 

The Pescadores islands consist, as Edmund Naumann kindly informs 
me, of alternating layers of basaltic tuff and coral limestone. 

While the north to south trend of Formosa might, perhaps, suggest 
a continuation of the northerly direction of north Luzon, the southern 
extremity of the great arc of the Lm-Kiu islands already deviates con- 
siderably from this direction. Doderlein has recognized the important fact 
that in the whole northern part of this arc, i.e. from about lat. 25** 8(/ N. 
to as far as Eiushiu, we may distinguish between an outer series of islands, 
turned towards the great Ocean, and an inner series. The outer series 
comprises the great non- volcanic islands of Okinawa-shima, Tukono^shima^ 
Amami-o-shima, Yakuno-shima, and Tanega-shima, while the inner chain 
is formed of smaller islands, which so far as they are known are of volcanic 
origin ; such are (perhaps) Kume-shima, then the volcano Tori-shima (Sul- 
phur island), then (probably) the Linschotens, Erabu-shima, Eo8e-shima» 
Tuo-shima, Tage-shima, and directly joining the arc we meet with Satsuma — 
Fujiyama on Eiushiu, the volcano Sakura*shima, which forms an island 
in the gulf of Eago-shima, Eiri-shimi-yama, and finally the famous 
Aso-yama in Eau shiu ^. 

Thus in the arc of the liu-Eiu islands that arrangement is repeated 
which we have already observed in the Antilles, the Nioobai; and Andaman 
islands, and the Banda islands, and which is so closely paralleled in the 
Carpathians and the Apennines. The fragments of the Cordillera stand 
on the exterior, the volcanos on the inner side. At the same time the 
extension of the arc into Eaushiu is unmistakable, and the south-east 
part of Eiushiu appears to be the continuation of the outer zone of the 
Liu-Eiu islands. 

In Oldna/wa Perry and Jones met north of Naf a with gneiss, day slates» 
and beds possibly containing coal : the most striking feature in the land^ 

^ Several travellers have described the rock as porphyry. 

* L. Doderlein, Die Lia-Kia-Insel Amami Oshima; Mitth. deatsch. Ges. Ostasiens, 
Yokohama, 1881, Heft 24, 31 pp., map, in particular p. 2. 



CH.ni] JAPAN 177 

fleape of the island, however, is formed by a steep ridge of -fossiliferoos 
limestone 400 to 600 feet high, which is broken up into a number of jagged 
peaks, probably by the same process of erosion as forms ' Earrenf elder ' 
(ur grykes ; it runs through a great part of the island, directed N. SO"" to 
6(f E. The gneiss is strongly folded \ 

Amami-ihshi/ma, according to Doderlein, is mountainous and consists 
of gneiss, granite, granulite, and crystalline schista Clay slates steeply 
upturned are to be seen on the east coast. 

8. Japan. The islands of Japan, together with New Zealand, are 

peculiarly well adapted to throw light on the structure of the wonderful 

island arcs which surround the Pacific Ocean, for they consist of fragments 

of simple Cordilleras ; the general plan is not complicated by virgation as 

in the Philippines, and fragmentation has not advanced so far as in the 

Banda islands and the arc of the Liu-Eiu islands. Fortunately we possess 

a recent account by Edmund Naumann. It is based on the result of many 

years of arduous work carried on with the co-operation of able Japanese 

geologists and contains a large number of fresh observations '. The author 

has veiy kindly furnished me, both in written and oral communications, 

with additional information on many points, so that it is on his observa- 

tioDS together with Milne's account of the volcanos of Japan ', and the 

much older map of Tezo by Lyman ^, that the following description is 

chiefly based. 

In the configuration of Japan a certain symmetry may be easily 
recognized. The great island of Hondo, bent in a gentle arc, is accompanied 
on the north-east as on l^e south-west by an extensive island — Yezo on 

* IL C. Peny, Narratiye of the Expedition of an American Squadron to the China Seas 
and Japan, compiled by F. L. Hawks, 4to, Washington, 1856, I, pp. 184, 811 ; G. Jones, 
Report on a Greological Exploration of Lew-Chaw, op. cit., II, pp. 53-56; and 
Ddderlein, op. cit, p. 27. £lie de Beaumont quotes fossils collected by the missionary 
P. Faret at Nafa in Okinawa, the age of which, however, cannot be regarded as certainly 
establifllied ; Compt. rend., 1859, XLVIII, p. 287, and J. Marcou, Lettres sur les roohes 
dn Jura, 8yo, Paris, 1857-1860, p. 269. The coral limestone, 200 feet above the sea and 
&r distant from it, is mentioned by R. H. Brunton, Notes taken on a visit to Okinawa 
Shimay Loochoo; Trans. Asiat. Soc. Japan, Yokohama, map, 1876, IV, pp. 66-77, in 
particalar p. 72. 

* S. Nanmann, Ueber den Ban und die Entstehongder japani8chenInseln,8vo, Berlin, 
1885, 91 pp. ; by the same, Die Erscheinnngen des Erdmagnetismus in ihrer Abhftngig- 
keii von der Erdiinde, 8vo, Stuttgart, 1887, in particular p. 15 et seq. ; Geological Survey 
of Japan, Beconnoisaance map. Geology, Division I, by E. Naumann, assisted by Takao 
Figitanai, Akira Yamada, Ichitaro Ban and Shogo Nishiyama (publication not yet 
complete). A little geological sketch*map of earlier date appears in J. G. Godfrey, 
Notes on the Geology of Japan ; Quart. Joum. Geol. Soc., 1878, XXXIY, pp. 542-554. 

* J. Milne, The Volcanos of Japan ; Trans. Seis. Soc, Japan, Yokohama, 1886, IX, 
pL 2, 184 pp., maps. 

* B. Smith Lyman, Geological Survey of Hokkaido, A geological Sketch-Map of the 
Island of Tesao, Japan, fol., Tokei, 1886. 

u N 



178 THE BORDERS OF THE PACIFIC OCEAN [paktiu 

the one hand and Eiushin on the other, and each of these advances a little 
further towards the Ocean than the adjacent end of Hondo. 

As we have already seen, the arc of the Lia-Kia islands extends from 
the south-Boath-weat into the island of Kiushia, so that some of Uie 
volcanoB of Einshiu must be regarded as a part of the ate of Ido-Kio. 



Yla. . 17. Japan. 

Tbedoeely eroea-hatohed lone trRveniiig the middle of Honshiu indiMtM the 'great lorn;' 

the Eone -wiUi aimple hatehing irhich mna through Sbiboku and KJi, showB the atrika of th* 

crystal] ine schista. Tbeutldixma of thewestooaat (Iwaki,MorioBhi, and others) are indieatad 

by lighter croBB-hatohing. 

At the same time the south-east part of Eiushin forms the continuation 
of the outer zone of ancient rocks of this arc, that ia of Okinawa-sbima, 
Amami-o-ahima, &c., and in fact this zone approaches in the island of 
Tanega-Bhima quite dose to the south coast of Eiushin. 

Similarly the volcanic arc of the Kuriles coming from the north-east 



CH.ra] SmKOKU AND KH 179 

enters into the body of Tezo. The connecting link is formed near Nemuro 
by the peninsula jntting ont of the eastern half of Yezo, and Milne 
mentions four volcanos in the eastern part of Yesso, two of them active, 
which must be indnded in the range of the Koriles \ 

The seas which snrronnd Japan, particularly the adjacent part of the 
Pacific Ocean, are of very great depth ; in fact tiie oceanic abysses sounded 
here are among the deepest known, while the island-strewn gulf which 
penetrates into the interior of Hondo has a depth of only 30 fathoma 
This gulf, therefore, is due to only a slight submergence, which has 
followed indeed the trend of the arc. 

The gulf or inland sea is separated from the Ocean on the south by the 
large island of Shikoku, and this is followed on the east by the peninsula 
of Eii which agrees with Shikoku in structure. The middle of the Pacific 
coast of Hondo is met by the chain of volcanic islands, which includes the 
island of OHshima; this we will call the Shishi-to chain. North of the 
bay of Tokio the coast trends more and more to the north, and bulges out 
in two broad projections separated by the bay of SendaL The southern 
of these regions which protrude out of the general course of the arc, we 
win call with Naumann the Abukuma mountains, that north of the bay of 
Sendai the Kitakami mountaina 

The holy mountain Fujiyama, with its symmetrical ash*<x>ne rising 
12,400 feet high, marks the region in which the volcanic chain of Shishi-to 
meete the middle of Hondo. A mighty line of division here cuts am»8 
Hondo from side to side; it is signalized by a great accumulation of 
volcanic products and is of the first importance in a study of the structure 
of the chain. Naumann speaks of it as the great region of fracture or the 
' great foss ' : it divides the island into two moieties, which we will call 
north Japan and south Japan. 

Shikoku and Eii, situated on the south-east coast of Hondo, are formed 
of stratified zones striking to the east-north-east; Radiolarian slates of 
Palaeozoic age, upper Carboniferous limestone, Trias, Jurassic, Cretaceous, 
and some Tertiary beds are present; they are thrown into folds, either 
normal or slightly overturned towards the Ocean, and form in a quite 
unmistakable maimer the outer zone of a mountain chain folded in the 
direction of the Ocean, that is to the south-east. Mesozoic deposits occur 
also as narrow bands pinched into the Palaeozoic complex. Behind these 
folded beds, that is towards the interior, a zone of crystalline schists very 
steeply upturned runs through Shikoku and Eii : characterized by great 
resistance, they stand out in the configuration of the country and form 
the long tongues or peninsulas which proceed from both sides of Shikoku 
as well as of Eii, and constrict the passage from the Ocean to the interior 

^ These are : Iwo-san, actiTe, with a caldron of boiling mud and sulphur, Ensori, 
Oakaa, with a regular conical form, and Meakan, an active cone* 

N 2 



180 THE BORDERS OP THE PACIFIC OCEAN [PABTin 

sea. From the western part of Shikoku in particular a long spur extendi 
towards the east of Eiushiu, and a glance at the map will show that in 
spite of the penetration of the lau-Eau arc into Eiushiu, the general strike 
of the folded ranges of south Hondo is also continued into a part of this 
island* 

The continuous course of this more resistant zone of schist, with its 
wing-like projections, gives to Shikoku the form of a double rhomb and to 
Eii of an irregular simple rhomb. I mention this fact because the same 
causes have produced the lozenge-like form of the isle of Wight, which is 
also due to the prominence of more resistant beds^ and because we shall 
meet with the same phenomenon on a larger scale in the outlines of Yeza 

The rocks of the outer zone of Shikoku and Eii are continued still 
further through Shima, which is formed by the eastern spur of Eii, but 
towards the great division in the middle of Hondo the strike bends from 
east-north-east to north-east and finally to north-north-east, and so describes 
an arc such as is usually formed by folds about to meet a second mountain 
range in syntaxis. Since this segment of the chain, bent in an arc, is cut 
off by the zone of division or fracture which strikes N. 25** W., a wedge 
results which Naumann calls the Akaishi sphenoid. 

On the inner side of the long schist range of Shikoku and Eii older 
rocks occur^ principally granite, which is weathered into a number of 
rounded summits and rises in innumerable islands out of the interior sea« 
The ranges lying towards the Japanese sea also consist of gFanite, ancient 
schists, quartz porphyry, and ancient massive rocks of various kinds. The 
north coast is distinguished by a few caldron-shaped subsidences, among 
which that of the volcano Daisen and the Sampei caldron may be mentioned 
in particular. The strike of these inner zones corresponds with that of 
Shikoku and Eii, with the same deflexion towards the region of fracture. 
On Eiushiu a cluster of recent volcanos in the middle of the island corre- 
sponds to the continuation of the interior sea. In the north-west of Eiu- 
shiu, Naumann believed he had seen a great granitic laocolite ; certainly 
a considerable part of the Japanese granites first made their appearance 
towards the dose of the Palaeozoic period. 

Naumann believes that it is the Chinese chains, striking out in the 
Tshu-san islands, which are continued through Eiushiu into the chain of 
south Hondo. A further examination of Eiushiu will show us in what 
way the arc of the liu-Eiu meets it, whether directly or in syntaxis. 

We have now reached the region of fracture. The line of the Shishi-to 
volcanos, coming from 0-shima, the last great eruption of which took place 
in January, 1877, meets the peninsula of Izu ; here rise Amagi-san (4,700 
feet), Hakone-yama (4,474 feet), Ashidaka-yama and the mighty Fuji-san. 
These are followed by a serried row of numerous other great volcanos, 
among them Tatsuga-taka (9,114 feet) with its twin peaks, Asama (8,800 






cH.m] ABUEUMA AND KITAKAMI 181 

feet) where an eruption oocnrred in 1870, Benge-san (9^00 feet) ¥dth two 
eratersy Tate-yama (9,400 feet) which stands on granite, and others \ 

On the other side of the great foes the stractnre of south Japan is 
repeated in its main lines. Hie Abukuma mountains and the Eitakami 
moontains consist of a series of f oesiliferous rocks similar to those of 
Shikoku and Kii, and similarly thrown into folds, which strike parallel to 
the general direction of this part of the island, that is fairly to the north. 
Towards the south, in the mountain region of Quanto, the folds are bent 
back towards the zone of fracture as in the Akaishi sphenoid on the other 
side of this zone. In this case, however, the bending back is accomplished 
by a change from north and south, first into a south-west and then finally 
into a north-westerly direction ; the north-westerly strike of the inner zones 
persists nearly to the middle of north Japan. The central region, however, 
which should correspond to the position of the interior sea^ is occupied by 
a long series of volcanos which we will term the meridional chain of north 
Japan. A group of five volcanos rises in the south towards the fractured 
region near the beginning of this chain ; the highest mountain of this group, 
Shipane-san (8,500 feet), was in eruption in 1872 ; Milne mentions 21 volcanos 
in Uie meridional chain. A part of western Yezo also belongs to it, with 
18 volcanos, many of them active. 

Finally the west coast of north Japan is characterized by four great 
caldron insinkings, each of which encloses a great volcano; these are, 
from south to north, Owassan, Chiokai (7,100 feet), Moriyoshi (5,800 feet), 
and Ounju-san (7,000 feet), which gives off a little steam ; its last erup- 
tion was in 1824. These caldrons which have been let down into the 
ancient rocks of the inner side of the chain are similar in every respect 
to the bays of Naples, Salerno, Santa Eufemia, and Algiers, and the other 
caldron inbreaks which surround the western Mediterranean on the inner 
Side both of the Apennines and of the Cordillera of north Africa. 

We thus see that south Japan consists of a cordillera of unilateral 
structure, folded towards the Ocean : its subsidences are represented on 
the inner side by the caldrons of the Sampei and Daisen; its inner zone 
appears in Chugoku and is marked by the granite rocks of the interior 
sea; its outer zones lie in eastern Eiushiu, in the mountainous country of 
Shikoku and Eii, and bend inwards in the Akaishi sphenoid to meet in 
syntaxis a second cordillera, that of north Japan. North Japan also is 
traversed by a cordillera ; the subsidences on its western inner border are 
marked by great caldrons from the Gwassan to the Iwaki ; its outer zones 
start from the syntaxis in the mountainous country of Quanto, describe an 
arc» and then proceed through the mountains of Abukuma and Kit.akami. 

^ Seyeral of the volcanos in the great fractured region have been described by B. von 
Draache in Bemerknngen fiber die japanischen Yulcane Aaama-Yama, Jaki-Yama, 
I^vawam-Tama nnd Fna-Yama ; Tbchermak*s Min. MittlL, 1877, pp. 49-60. 



182 THE BORDERS OP THE PACIFIC OCEAN [PABrra 

The syntaxis indicated by the inflexion towards the interior exhibited 
both by the Akaishi and the Quanto region, but in opposite directions, is 
distinguished from all other cases of syntaxis by the fact that a lai^e part 
of the syntactic region has subsequently collapsed and great volcanoB have 
arisen in the trough, on the continuation of the Shishi-to chain. 

There is some difference of opinion, not as to the facts but as to the 
precise manner of interpretation. Naumann prefers to regard the great 
trough not as a recent subsidence but as a very ancient fissure, although 
not so old^ as the main longitudinal fracture of the whole range, and he 
assumes the presence of a pre-existing obstacle to the folding, somewhere 
in the position of the Shishi-to chain. Harada, however, adopts the same 
explanation as we have given above K 

Milne has incidentally made a comparison between the Shishi-to chain 
and the Ladrones; this would involve the existence of another in<fnVi» 
arc, syntactic with the arc of south Japan, and syntactic against that of 
north Japan. But such a supposition does not accord with the existence 
of the downthrown syntactic fragments encountered within the fractured 
region; and still less, with the aspect presented by other regions of 
syntwda. 

Our knowledge of Yezo is unfortunately less complete than that of 
Hondo. Pumpelly, who some time ago examined the south part of the 
island, encountered on the southern peninsula, between Hakodate and 
Volcano bay, a range of ancient schists with quartz porphyry and green- 
stone which strikes to the north-west and must probably be regarded a? 
the continuation of the northern arc of Hondo. 

This band is surrounded by more recent volcanos and by recent marine 
deposits arranged in terraces K From Lyman's map we may conclude that 
a long range of folded rocks strikes across the island through its greatest 
breadth towards the north-north-west in the direction of Saghalin. The 
island owes its lozenge-like form to this great range, thus recalling 
Shikoku, Kn, and the isle of Wight. We must perhaps regard it as the 
continuation of Saghalin. It has also been represented as the continuation 
of the northern arc of Hondo carried further towards the Ocean, but the 
fragment striking to the north-west, which Pumpelly observed near 
Hakodate, does not support this conjecture. We also know from Naumann 
that in central Yezo the middle Cretaceous covers a large area and is 
probably directly superposed on ancient rocks. In like manner, according 
to the investigations of F. Schmidt and P. von Glehn, the Cretaceous in 
the south part of Saghalin attains a very wide development, and the 

^ Nanmaim, Erdmagnetismos, p. 18; Harada-Toyokitsi, letter in Anz. k. Ak. Wias. 
Wien, July 7, 1887. 

' R. Pumpelly, Geological Beflearches in China, Mongolia and Japan ; Smithsoaiaa 
Contributions, No. 202, 1866, p. 79 at seq., in particular p. 106, pi. nii. 



CH.ra] THE KURILES 183 

impression made on Naomann also was that this Cretaceous of Saghalin 
would be found to traverse the whole of Yezo or would certainly be 
eonttnoed into the peninsula. Pending further information, however, it 
would be well not to directly connect Saghalin and the principal chains 
of Yezo with the principal chain of Hondo, but to regard them as forming 
an independent feature lying further to the east and distinguished by 
a eoosiderable development of the Cretaceous \ 

In conclusion we know that a part of the east of Yezo belongs to the 
▼olcanic arc of the Euriles. 

9. The Kuriles and Kamchatka. The Euriles have been visited three 
times by Milne. They contain twenty-three well-formed cones, sixteen of 
them smoking^ arranged in a long arc. Sedimentary beds or older rocks 
have not becod encountered; the elongated islands, such as Iturup, 216 
kilometers long, appear to have been formed by the action of marine 
earrents which, flowing parallel to the arc. have united the cones by piUng 
up the ashes between them \ 

The island of Paramushir, more than 90 kilometers long, consists of 
a series of extinct volcanos, with one crater still smoking, directed to the 
north-east and surrounded by an accumulation of ashes and lavas. Only 
a narrow arm of the sea separates this island from the flat island of 
Shomochu, which follows on the north-east and also consists of ashes and 
lavas ; this in its turn is only separated from cape Zopatka, the southern 
point of Kamchatka, by a channel of no great depth, due to marine 
erosion. North-west of Paramushir there rises out of the sea the steep 
isolated cone of Alaid, which last broke into eruption in 1793. These are 
the most northern of the Euriles, and now the great train of volcanos, 
which we have been able to trace from the east of Yezo, enters the 
peninsula of Eamchatka. The series proceeds through the eastern part 
of the peninsula, comprising according to Dittmar thirty-three volcanos, 
twelve of them active \ The most northerly of these are the Eljutshewska 

* For Saghalin : F. Schmidt, Sachalin ; Baer and Helmersen, Beitr&ge zur KenntniBS 
des rmBischen Reiches, 1868, XXV, p. 177 et seq. ; P. von Glehn, Beisebericht von der 
buel Sachalin, torn, cit., p. 189 et seq., in particular pp. 203-277 ; also Schebunin's 
map of Sachalin; further, F. Schmidt, Ueber die Petre&cten der Insel Sachalin, 
M^m. Ac. Imp. Saint-P^tereb., 1873, 7^ s^r., XIX, No. 3. For Tezo : E. Naomann, Ueber 
das Vorkommen der Kreideformation auf der Insel Tezo (Hokkaido), Mitth. deutsch. 
€tes. Ostasiens, Yokohama, 1880, Heft 21 ; and by the same, Ban nnd Entstehung der 
japanischen Inseln, p. 21. 

' Milne, A Cruise among the Volcanos of the Kurile Islands ; Geol. Mag., 1879, 2nd ser., 
yi, pp. 337-348, map ; and Volcanos of Japan, pp. 125-169, map. 

' C. Ton Bittmar, Die Vulcane Kamtschatkas, Peterm. Geogr. Mitth., 1860, II, p. 66 ; 
also A. Postels Bemerkungen fiber die Vulcane der Halbinsel Eamtschatka, M^m. Ac. 
Imp. Saint-Petersb., 1835, II, pp. 11-28, and Perrey, Document sur les tremblements de 
teire et les ph^nom^es volcaniques dans Parchipel des Kouriles et au Eamtschatka, 
166 pp., Sto, 1863 (in the Ann. Soc. agric. Lyon). 



184 THE BOEDEES OP THE PACIFIC OCEAN [PAJirm 

Sopka (lat. 56" 8' N., 4,804 meters) and the Shewelutsh (lat 56* AKf N., 
8,216 meters) ; both these great mountains have been described in detail 
by Erman^ In the axis of the peninsula a broad longitudinal valley^ 
stretching to the north-east, corresponds to the upper course of the river 
Kamchatka; this river after following the valley for a great part of its 
length turns almost at right angles, and flowing to the east at the north 
foot of the Eljutshewska Sopka reaches the sea near Nishne Kamchatka 
The series of active volcanos extends east of this valley to the Kljutshewska 
Sopka, and the Shewelutsh alone rises north of the lower course of the river. 

Dittmar has published a general geological map of the peninsula^ 
in which he has incorporated Erman's observations ^ Judging by the 
structure of the Eiu-Liu, the Antilles, and other island arcs, we might 
expect the subsided cordillera of the Kuriles to become visible in 
Kamchatka on the east side of the volcanoa Dittmar does in &ei 
mention ancient crystalline schists and similar rocks on all the peninsulas 
of the east coast to beyond lat. 58"" N., with the exception of the broad cape 
E[ronozkij, which is covered with ashes and lavas; granite and gneiss 
appear here and there at the very foot of the volcanos. Analogy would 
certainly lead us to regard these rocks in the east and those at the foot 
of the volcanos as the few visible remains of the cordillera of the Karile& 
Many ranges of granite, gneiss, and ancient schists of greater importance 
appear, however, to the west of the great volcanos, and in particular a long 
range lying west of the upper course of the Kamchatka river; this is 
apparently 2f degrees of latitude in length and accompanied by day slates; 
to the north follow mountains of porphyry. These again are joined by 
recent volcanic ejections; between lat. 56° 10' N. and lat. ^V9f^ N. there 
rises in the west a group of five extinct volcanos forming an independoit 
western zone. 

Two fragments of arc would thus appear to occur in Kamchatka — an 
eastern segment with the great active volcanos and the ancient rocks of 
the eastern promontories, representing the continuation of the Kuriles, and 
a western segment to which would belong the range of hills west of the 
Kamchatka valley, tc^ether with its prolongations and the extinct volcanos 
of the north-west ; this is just what we find in Tezo, where indeed we 
were able to distinguish three successive arcs, one behind the other. 

^ A. Erman has already published a geological map which embraces Kamchatka in 
Erman's Archiv ftU: wissenschaftliche Eunde von Russland, 1842, II ; in the foUowiiig 
Toltune there is a description of some fossils from Kamchatka bj Girard. 

' Dittmar, Einpaar erlftutemde Worte zur geognostischen Karte Kamtschatkas ; Boll 
Ac. Imp. Saint-P^tersb., 1855, XIV, pp. 241-250, map. Professor KrentE, of Cxaeow, 
has kindly sent me a list of the rocks collected by Professor Dybowski in Tra.iwAiiftf ,]ra and 
the Commander islands, and observes among them, from Tigil (KamchatkaX light giey 
compact siliceoas limestone filled with Trochammiha, Haplophragmiom, T^^^na and 
other rhiysopods. 



CH,m] EEVIEW OP THE ISLAND ARCS 185 

The west coast in the neighhoarhood of Bolaherezh, and even much 
farther to the north, is flat. In the west and in the K'fLTnf>lm.^lrfl. valley 
Tertiaiy coal beds with leaf imprints make their appearance. They also 
occur far to the north, on cape Tajganos and northwards as far as Ishiga, 
lying in patches on the granite and the ancient schists of the peninsula. 

10. Qeneral survey of the island arcs. It follows from this investiga- 
tion that in the north-east of Asia the following arcs may be distinguished: — 

(a) The arc of the IdvrKiu islands. This consists of the fragments of 
the Cordillera which form an outer zone, and of the volcanic zone on the 
inner aide ; it extends into the south of Eiushiu. 

(b) The arc of scmtk Japan. Its outer zone extends from Eiushiu 
through Shikokii and Elii, and is bent backwards in the Akaishi sphenoid 
in the direction of a syntaxis against the great fractured region which 
distinguishes the middle of Hondo. Naumann supposes that it is continued 
to the folded ranges of south China which run out to sea near the Tshu-san 



(c) The arc of north Japan. The outer zones proceed from the fractured 
region, bending through the mountains of Quanto to continue through 
Abukuma and Eitakami. The inner zones bear the volcanic meridional 
chain and the west coast is characterized by caldron inbreaka A part of 
the continuation of the arc directed to the north-west lies in the southern 
part of Yezo. 

(d) The arc of centraZ Yezo and Saghalin, situated somewhat nearer 
the Ocean, is distinguished by the development of the Cretaceous system. 

(e) The arc of the Kuriles. The east of Yezo belongs to the arc which 
runs into southern and eastern Kamchatka as far as the volcano of 
Shewelutsh. It is represented as far as Kamchatka by volcanos only, 
but no doubt the zone of older rocks lying east of the Kamchatka river 
together with the volcanos may be regarded as a part of its cordillera. 

(/) The fragment in central and western Kamichatka. 

11. 2%6 n/yrth of China. The question now arises as to the relations of 
these island arcs with the mountain chains of the Asiatic continent. We 
were able to trace the Malay arc from Yunnan to the west coast of New 
Guinea. Sut the chains of the Philippines diverging towards the south 
presented no visible continuation on the mainland. The trend-lines of the 
northern arcs and the adjacent* ranges of China have often been repre- 
sented diagrammatically, but the sdiiemes even of the most competent 
authorities, such as Pumpelly and If(5czi, differ completely from one 
another ^ 

As the result of a laborious journey from the central Yang-tse-kiang 
through Peking into Mongolia, Pumpelly was able to confirm the statement 

' IHimpellj, Geological Researches, &c., pi. vii ; L. Ldczi, A Khinai birodalom term^ 
seti ViBonyaiiiak ^ Onz^lgaiiiak Leidua, 8vo, Budapest, 1886, p. 19. 



186 THE BORDEBS OF THE PACIFIC OCEAN [pabthi 

of earlier travellers that the mountain chains of China strike chiefly from 
south-west to north-east (more exactly W. SO"" S. to K 80"* N.), and he 
distinguished the mountains which follow this direction as the Sinian 
sydem \ But it will only be possible to obtain a general idea of this vast 
empire when Bichthofen's comprehensive works have been completely 
published ^. In the great work with which F. von Bichthof en has enriched 
our science he has not only recorded observations accumulated daring 
many years of travel, but has also attempted to analyse the structure of 
the land into its various elements and to determine the relations of these 
elements to one another. In the south, long folded ranges predominate, 
some striking out to the sea in the Sinian direction, ie. to the north-east, 
the others in the south-west running from Ta-tshon-fu and further thromgh 
Yunnan to the south-south-west, and as it now appears forming the 
beginning of the Malay arc. In the north, however, the structure of 
the country is much more difficult to interpret ; particularly so north of 
Han-tshang-fu, where near lat. 33** N. the Sinian chains abut against the 
south side of the rectilinear chain of the Tsin-ling-shan, which trends 
east-south-east, and thence into the southern part of Mongolia. Fragments 
of ancient table-land are present, and folded ranges of different age and 
direction, as well as subsidences, also of different age. Bichthofen has 
distinguished and defined these regions; master of a vast field of fresh 
observations, he has arrived at general conceptions which accord in the 
most happy manner with views on the origin of mountains simultaneously 
developed in other landa Thus, what is described in China as a ' diagonal 
mountain,' i. e. as a mountain in which the direction of the folding does not 
correspond with the orographical outline, is in fact a horst, the visible 
remains of an ancient folded range. In a number of very important 
points the description of China has confirmed views already accepted in 
Europe. This is particularly true of Bichthofen's statement, that in a 
country already folded fresh folds or even faults may be produced which 
follow the original direction K 

No marine deposit of Mesozoic or Tertiary age is known anywhere in 
the north of China or in the south as far as Yunnan. Tha sediments of 
the lower or middle Jurassic contain only terrestrial plants and coal bed& 
They are directly succeeded by lacustrine deposits of very recent date, 
then by the loess and the alluvium of the great plain of the Hwang-ho. 
Even the Silurian and Devonian sediments have not yet been encountered 
between the Weiho and Mongolia, that is in the whole of China north of 

1 Pampelly, op. cit., p. 67 et aeq. 

* F. Ton Richthofen, China, ErgebniBse eigener Reisen and daranf gegrdndeter 
Studien, 8vo, Berlin, I, 1877 ; II, Das nOrdliche China, 1882 ; IV, P^l&ontologischer 
TheU, 1883 ; Atlae, fol. 1, 1885. 

' Bichthofen, op. cit, II, p. 687. 



CH. m] CHINA 187 

the Tsin-Iing-ahan; in that region the Carboniferous lies immediately on 
the Cambrian beds, and is followed by the Mesozoic. 

The seyeral elements of the structore do not stand out very obviouslyi 
since a part of the interval between them is occupied by the shallow 
waters of the gulf of Pe-chi-li, an inland process of the Yellow sea, and 
another part is concealed by the alluvial land of the Hwang*ho. If we 
imagine this alluvial land to be submerged as far as the foot of the 
mountains, that is, to Peking, Hwai-king-fu, and Nanking, then the 
mountain mass of Shjan-tufng will rise out of the waters as an island. 

This mountain mass is cut right across by a great fault along the 
Wei-ho. The western half shows the Archaean foundation covered by 
horizontal Cambrian and Carboniferous beds, and it sinks to the north 
al<»)g &ults running in the same direction but not parallel In east 
Shan-tung the Archaean foundation is laid bare ; a folded chain trending 
to the north-east once existed here ; it was denuded before the Cambrian 
period, and on its planed-down surface rests a horizontally stratified mass 
of Cambrian deposits ^ Thus in Archaean times folding already existed 
here, and the strike of these primaeval folds does not differ materially 
from that of the recent folds in south Japan. Since the Cambrian period 
no folding has taken place in this region ; fractures alone have occurred. 
West Shan-tung appears to be the downthrown part of a shattered horst, 
since here the sedimentary covering is more completely preserved. The 
fractures in west Shan-tung are, in part at least, of pre-Carboniferous age, 
as is shown by the transgressive bedding. 

Liao-iuTig has the same structure as east Shan-tung and may be 
r^arded as its continuation. Here, too, the fundamental rocks are folded, 
and we meet with superimposed patches of horizontal Cambrian beds ; on 
the border of Korea these have furnished Trilobites^ 

The journeys subsequently made in Korea by Gottsche show that these 
deposits extend into north Korea; that by far the greater part of this 
peninsula consists of crystalline schists, and that, with the exception of 
tiie Tertiary lignite beds, the small patches of recent strata overlying them 
indade no sediments of later date than those of north-east China ^. 

' Bachthofen mentions at several places a still older direction which is said to be 
recognisable in the most ancient gneiss ; NE.-SW. is the strike of the folded mica-schists, 
primitiTe limestone, ftc. 

* Dames compares two localities with the lowest division of the American Potsdam' 
sandstone ; the third, from the presence of the genus Dorypyge, most closely corresponds 
te the Quebec group of Utah. He leaves it an open question whether this horizon 
should be assigned to the lower Silurian, or, as Barrande thinks, to the Cambrian; 
W. Dames, Gambriache Trilobiten von Liautung, in Richthofen, China, lY, p. 33. 

' C. Gk>ttBche, Geologische Sldzze von Korea ; Sitzungsber. k. preuss. Akad. Wiss. 
Berlin, XXXYI, Sitznng vom 15. Juli 1886, 17 pp., map. The island of Mackau in the 
archipelago lying south-west of Korea has been visited by Guppy ; it consists of gneiss, 



188 THE BORDEES OF THE PACIFIC OCEAN [PARrm 

Thus on either side of the Yellow sea stands a primitive orographie 
mass, Korea with liao-tung on the one hand and Shan-tung on the otlier ; 
each is characterized by the fact that Cambrian sediments rest in 
horizontal sheets on an Azoic foundation. The structure of these r^;ioii8 
is whoUy different from that of the neighbouring Japan. 

The accounts hitherto published do not enable us to speak with equal 
certainty of Liao-hsi, but pending further information, I have no hesitation 
in regarding the greater part of it as belonging to the next mountain 
fragment. 

A region composed of parallel mountain chains has been graphicaDy 
described by Richthofen as a grill or grid (^Bosf), and in this sense we 
speak of the grill of Peking. Its ranges strike E. 30* N. How fax it is 
continued past Tshong-to-fu in the same direction into eastern Mongolia 
is at present unknown. To the north it is in great part covered by leoent 
lavas, which appear outside the Great Wall in broad platforms, but beyond 
these platforms and even projecting from them fragmente of the griU «» 
still to be seen. To the west the grill ends suddenly against the plains of 
Ta-tung-fu and Hsin-tshou, along a line which cuts the strike of the ranges 
transversely. The direction which is distinctive of the grill dates from 
a very early period, for while the gentle undulations of the gneiss maintain 
the characteristic direction, e. g. in the lofty mountain of Wu-tai-shan, the 
Cambrian beds lie flat. But of much greater importance, as regards the 
existing form of the grill, than these ancient folds appear to be the flexures 
and longitudinal fractures with dragged-back edges which follow the 
original direction, and throw down the whole region in successive steps 
from Mongolia towards the great plain. 

These subsidences possibly date from different periods; in any case 
they are much more recent than the pre-Cambrian folds. 

The mightiest and longest range of the grill begins in the south-west 
with the gneiss dome of Wu-tai-Juin, which is over 10,000 feet in height 
and lies north of Hsin-tshou; this is continued to the north-east in the 
Hsian-wu-tai-shan, and then in the Nankou chain, which sinks in the south- 
west by a great flexure to the plain of Peking. The range of Wu-tai-shan 
is followed on the south by that of Hsi-tshou-shan, which is doubtless 
continued to the north-east in the granite ridge of Hong-shan« The latter, 
instead of proceeding to the north-east like the others, coUapses/and it is 
only on the other side of the plain of Peking that the Pan-shan rises in the 
same line of strike. This sunken area of Peking is thus open to the south- 
east towards the great plain, and bounded on the north-west by the flexure 
of Nankou ; it is in fact an inbreak between Hong-shan and Pan-shan. 

Nearly six degrees of latitude south of Peking lies Hsi-ngan-fu, and 

greisen, and quartzite ; Notes on the Geology of the Cotean Archipelago, Nature, March 
8, 1881, pp. 417, 418. 



CH.in] TSIN-LING^SHAN 189 

9onth of this great town rise the snow-covered peaks of the mighty TevTir 
Hng-ahanj which trends in a straight line from W. 12^ N. to EL 12'' S. and 
marks the boundary between north and south China. It is a long range of 
unilateral stractore ; on its north side the gneiss reaches a height of over 
11,000 feet ; the more recent sediments lie to the south. In the east, on 
the road between Eai*fong-fu and Hsiang-yang-fu, it is completely faulted 
down across the strike; beyond this line it again emerges as the Hwai 
mountains, and attains a height of 4,000 feet. The last outposts may be 
seen near Nan-king, in the region at the mouth of the Yang-tse-kiang. 

The rectilinear east-south-east trend of the Tsin-ling-«han and its 
prolongations thus differs considerably from the east-north-easterly or 
north-easterly strike of the Peking district. But before speaking of this 
great range reference must be made to some regions situated to the north 
of it. 

The Wei-ho follows in its upper course the east-south-east trend of the 
Tsin-ling-shan and flows along the north foot of these mountains. A little 
above Hsi-ngan-fu, however, it abandons both this direction and the foot 
of Hie mountains and turns to the east-north-east; it thus reaches near 
Tong-hwan-ting the Tellow river, which here turns in a sharp elbow and 
then continues the direction of the Wei-ho up to the eastern termination 
of the mountains. In the triangular space confined between the river 
valley running to the east-north-east and the Tsin-ling-shan trending to 
the east-south-east (Aher ridges rise which in an orographical sense form 
the northern outposts of the Tsiii-ling-shan. Of these chains we must 
first separate off Hwa-shan, which trends in a different direction, and is 
situated dose to the confluence of the Wei-ho with the Hwang-ho; we 
shall refer to this later. The other ranges, chief among which are Fu-niu- 
dian near the base of the triangle and Sung-shan near its northern apex, 
abo follow an east-south-east direction, but their structure is not 
determined like that of the Tsin-ling-shan by folding, but by parallel 
fractures. We have here, indeed, the fragments of an ancient table-land 
formed of Cambrian sediments with directly superposed Coal-measures, 
bat without the Carboniferous limestone, elsewhere so widely distributed. 
The transgressive Coal-measures are thus situated behind the Tsin-ling- 
shan, folded to the south, very much in the same way as the Coal-measures 
of Bohexhia lie behind the Variscan chains, folded to the north. 

The Tsin-ling-shan has been crossed in various places by David ^ 
Bichthofen, Sz^chdnyi, and L6czi. The transverse section of which we 
have the most precise knowledge is that on the Tsin-ling road which, leads 
from the north into the basin of Han-tshung-f u ; it has been described in 
detail by Bichthofen. The northern slope is steep, the southern likewise, 

' A. David, Journal de mon troisifeme voyage d'exploiation dans TEmpire chinois, 
2 vols., 8to, Paris, 1875. 



190 THE BORDEKS OF THE PACIFIC OCEAN [pabthi 

and the range thus resembles a solid beam ; nevertheless, situated almost 
in the middle of this well-defined orographic r^on, lies an importaat 
tectonic boundary. Richthofen coming from the north first encountered 
a great band of gneiss accompanied by red granite, then Azoic schists and 
massive rocks, the Wutai-beds, and with these, rocks containing chlorite 
and hornblende. On this zone, patches of Carboniferous beds with 
anthracite are superposed in transgression, as is so often the case in the 
older folded ranges of Europe. The Wutai zone is followed on the soutii 
by folded beds, among which middle Silurian, upper Silurian, Devonian, 
and Carboniferous limestone may be recognized by their fossils. The first 
three members of this series are, as we have already stated, unknown in the 
whole of the vast region extending north of Tsin-lingnahan into Mongolia, 
where the Carboniferous limestone always rests directly on Cambrian beds ; 
and it is very remarkable to observe that here, as in the Alps, the Andes, 
and so many other mountain ranges, the series of marine deposits becomes 
complete as it leaves the table-land to enter the great folded chains. 

South of this folded zone a more recent granite mass emerges near 
liu-pa-ting. It is situated south of the middle of the range. A band of 
micaceous slate and sericite phyllites, perhaps altered Carboniferous, 
follows the mountains with a completely abnormal strike ; we then reach 
a zone, twenty-two geographical miles broad, occupying a third of the whole 
breadth of the chain, and with a completely different strike ; from the north 
border of the mountains up to the granite of liu-pa-ting the direction 
E. 12^ S., or the normal direction of the Tsin-ling-shan, prevails, but in 
this broad zone the strike is to the east-north-east. We have now reached 
the region of the Sinian direction of folding; in Sze-tshwan it becomes 
north-east and dominates the whole of south China. 

The rock with the Sinian strike which forms the broad zone on the 
south side of the Tsin-ling-shan is formed of micaceous gneiss, intensely 
folded with frequently repeated intercalations of crystalline limestone; 
perhaps, as Richthofen conjectures, Silurian sediments which have suffered 
metamorphosis. 

Before discussing the adjacent folded ranges of south China, let us 
return once more to the north. 

We have seen how the grill of Peking breaks off to the west against 
the plains of Ta-tung-fu and Hsin-tshou; the Hsi-tshou-shan forms its 
south-western extremity. From the border of this mountain fragment 
down to the Hwang-ho, a steep diff, which- forms the western boundary 
of the plain, runs past Pau-ting-fu and Tshong-ting-fu. * From Pau-ting-fu 
its direction is south-south-west, and further to the south before reaching 
the Hwang-ho it bends in a curve from south-south-west to south-west 
and west-south-west. This cliff, about 2,000 feet high, bears for its whole 
length the name TairJiang-shan, In the north, near Tshong-ting-fu, an 



CH. m] TAI-HANG-SHAN 191 

old^ mountain segment is visible; with this exception Tai-hang-shan 
forms the border of a vast table-land of Carboniferous sediments which 
extend from the north with horizontal bedding through the whole of 
aouthem Shansi and northern Shensi as far as the north foot of the Tsin- 
ling-fihan. A first stage, which consists of Carboniferous limestone and 
eoal-bearing sediments, is followed by a second formed of a supra-Carbo- 
niferous sandstone, and thus the table*land attains a height of 5,000 feet 
On the east and south-east it descends in steps to the great plain, and 
Tai-hang-shan itself is partly a flexure and partly a step fault. 

The way in which the southern part of Tai-hang-shan is bent back- 
wards is very striking, and strange to say this course is repeated in a long 
nanow range formed of gneiss and other ancient rocks which emerges 
from the Carboniferous plateau. This range bears successively the names 
Hwa-ahan (south of the Hwang-ho), Fong-tiau-shan, Hsian-mieuHsihan, and 
Ho-shan. I should not be averse, while awaiting further information, to 
regard this whole range as a horst left in relief amid the broad simken 
table-land. Beyond it flexures and faults running in the same direction 
follow one another up to the plain of Ta-tung-fu. This is succeeded on 
the west by a table-land presenting the same structure formed of a Mesozoic 
coal series. 

Orographic blocks let down in steps thus surround the great plain on 
the whole of its northern side. The flexure of Nan-kou above Peking and 
Tai-hang-shan^ the boundary face of the table-land, are the features most 
characteristic of this structure. 

In the south the case is entirely different. We have already become 
acquainted with the Tsin-ling-shan as an asymmetric range folded to the 
south, like all the great chains of Asia» and we have already pointed out 
that other folds join on to its south side which strike to the east-north-east 
and farther south to the north-east. The micaceous gneiss with crystalline 
limestone, which we mentioned as occurring on the south side of the Tsin- 
ling-shan, is followed south of the depression of Han-tshung-fu by a zone 
of folds formed of Silurian, Devonian, and Carboniferous limestone, and 
inverted to the south ; these folds are covered towards the south by a series 
of sediments lying unconformaUy. This new series begins with limestone 
of unknown age ; upon it rest coal-measures of the lower Jurassic ; these 
are covered by red, argillaceous, and sandy sediments, most likely of lower 
Jurassic age, which fill the ' red basin ' of Sze-tshwan. This more recent 
series has also experienced lateral movement and longitudinal fracture. 

The folds which meet the southern slope of the Tsin-ling-shan are 
continued, as we have said, through the whole of the south of China. 
For a distance of almost ten degrees of latitude, says Richthofen, the 
numerous parallel and closely crowded mountain ranges striking from 
west-south-west to east-north-east are cut off transversely or diagonally by 



192 THE BORDEES OF THE PACIFIC OCEAN [PAKrm 

the curve of the coast. We have seen that the same direction of folding 
also prevails in Tongking^ 

As regards the relations of China with the island arcs, the following^ 
must be borne in mind. 

In the neighbourhood of Nan*king the continuation of Tsin-ling-shan 
reaches the sea. North of this point the whole country consists of ancient 
table-land ; beneath it, pre-Cambrian folding of the Archaean and Azoic 
rocks, with a north-easterly strike, may be observed in places. The only 
marine deposits which have been recognized are those of the Cambrian 
system and the Carboniferous limestona The Tsin-ling-shan does not 
correspond in its course with the island arcs. To the south of it lie chains 
folded to the south in which middle and upper Silurian as well as Devonian 
are known, but no recent marine deposits on this side of Yunnan. The 
strike of the chains certainly coincides with that of south Japan, and here, 
according to Naumann, we must look for the continuation of the arc of 
that island '. 

12. The north-east of Ada. In Europe the recurrence of folding 
directed to the north, from the pre-Devonian overthrusts of north Scotland 
to the most recent movements of the Alps, has shown how extraordinarily 
constant in direction the tangential movement may be over an extensive 
area ; China affords a similar example. 

The pre-Cambrian folds of northern China, which were worn down and 
then covered by the flat-lying Cambrian sediments, present the same norUi- 
east strike as prevails in that vast region of far more recent folding which 
extends south of the Tsin-ling-shan down to Tongking. 

• This same north-easterly direction, characteristic of the Sinian system 
of Fumpelly, also prevails with certain deviations to the north-north-east, in 
the whole of north-eastern Asia' from the Great Wall to the Arctic Ocean. 
A study of the relations of this east Asiatic ridge trending to north-east 
and north-north-east with the Altai and the chains of Thian-shan, would 
require first of all a discussion of the extensive and laborious investigations 
which have been carried out by Russian geologists in the region of lake 
Baikal ; but such a discussion would lead us too far from the subject of 
this chapter, which is devoted to the outline of the Pacific Ocean, and it 
must, therefore, be reserved for a later page. For the present it will sufiEice 
to mention some of the principal features. 

Even in liau-hsi some indications, as in the T-wu-lu-shan, point to the 
occurrence of a strike directed to the north-north-east, but it is precisely 

1 Bichthofen, Ftihrer fiir Fonchungareisende, 8vo, Berlin, 1886, p. 309, 810. 

' Tahe-kiang and Fa-kiang are still unfortunately almost unknown ; Basset-Smith 
encountered granitic and felsitic rocks and steeply upturned schists on the islands of this 
coast ; see Notes on the Geology of part of the Eastern Coast of China, Nature, June 16, 
1887, pp. 168, 164. 



cH.ra] THE STANOVOI 193 

here that observations become lees connected. A fact of greater impor-^ 
taace is the existence of a zone of volcanic rocks running from south-south- 
west to north-norUi-east, which Bichthofen has traced from Wei-hsien in 
Shantung, past Tong-tehou-fu and the Miau-tau islands, through the valley 
of the liau; and he believes, it may extend into the neighbourhood of 
Meigen^. Bichthofen also points out that the Great Ehingan is not a 
mountain chain but the edge of a table*land, and that it nearly coincides 
with the prolongation of the edge of the Carboniferous table-land of Shansi, 
i. e. to the Tai-hang-shan ^. 

If from Peking we cross the eastern Gobi by the route described by 
Mufichketow, to the north-north-west, we encounter beyond the desert, on 
tiie right bank of the Sselenga, a grill of parallel chains which strike from 
south-south-west to north-north-east \ Wenjukow enumerates ten parallel 
chains besides the long range of the Apple mountains or Jablonowyi^. 
The Jablonowyi ridge represents, like the EJiingan, the steep face of 
a flexure dropped on the east, or a fracture; this is confirmed by all 
descriptions of the road from lake Baikal to Tshita, and by Eropotkin's 
express statement. Here begins the great ridge of the divide, the StaTiovoi, 
explored by Middendorf ; it is a plateau of variable breadth surmounted 
by secondary hill ranges; its east border trends to north-east or north- 
north-east across the valley of Amur to the sea of Okhotsk, of which it 
forms the west shore, and is then continued to the Arctic Ocean. On the 
east it is followed by successive chains, all striking to north-east or north- 
north-east ; among them, in particular, Dousse Alin, or Middendorf 's Bureja 
mountains, and Sikhota-Alin, the coast range of Manchuria. Kropotkin 
has published an extremely instructive sketch-map of the trend-lines of 
these chains, which exhibit throughout a certain tendency to converge 
towards the southern half of the sea of Okhotsk, and all disappear along 
the south coast of this sea^ Hence the great number of islanck and bays 
on this part of the coast, which is a true rias coast, like that of the south- 
east of China. In one of these bays, the bay of Tugur, west of the island 
of Elein-Shantar, Middendorf discovered on cape Earaulnoi, in Mamga 
harbour, contorted beds of clay slate, containing the first specimens of 

' Bichthofen, China, II, p. 50 et passim. The volcano Pei-shan at the sonrces of the 
Sungari appears to lie somewhat east of this line ; cf. Proc. Geogr. Soc., London, 1886, 
Yin, p. 779. Bichthofen refers here to the eruption of Ujnn-Choldongi of 1721 to 1722 
mentioned by Wenjnkow ; this region lies twenty-five versts SE. of Mergen on the river 
Nemer, not far from the mountain Dousse-Alin to be mentioned directly. 

* Bichthofen, op. cit, II, p. 520 et passim. 

' J. Mnschketow, Geologische Notizen fiber die Ost-Mongolei ; Gomoi Joum., 1881, II, 
pp. 80-98, with a geological map of the route from Dolon-nor to lake Tarei. 

* Wenjukow, Die russisch-asiatischen Grenzlande ; German translation by Krahmer, 
8vo, Leipzig, 1874, p. 187. 

* Kropotkin in £• B^cluz, Nouvelle G^ographie universelle, VI, 8vo, Paris, 1861, p. 813. 

lOIXAS u O 



194 THE BORDERS OF THE PACIFIC OCEAN [parthi 

PaeadomoTVotis Ochotica, which proves that the Trias was involved in the 
elevation of these chains. This fossil has acquired, as we shall see later, 
a peculiar importance in the history of the whole Pacific Ocean \ 

The furrow of lake Chanka, the Ussuri, and the lower Amur, as seen on 
the map, correspond very nearly with the west border of the coast range 
of Sikhota-Alin, and B^luz, in his description of this region, based on the 
observaticms of Erc^tkin, makes the just remark that a slight submer- 
gence would suffice to convert Sikhota- Alin into a new island arc ^ 

A comparison of Bichthofen's results in northern China with those of 
Russian geologists in east Siberia thus enables us to recognize the Tai-pai- 
shan in Shansi, the Great Khingan in Mongolia, the Apple mountains in 
the trans-Baikal region, and the east of part of the dividing ridge of 
Stanovoi from the Apple mountains to beyond Okhotsk, as the edges 
of great platforms which have been let down in the direction of the Pacific 
Ocean. * Certain facts,' says Bichthofen, ' indicate that the Bacific basin is 
surrounded by step faults on the grandest scale V 

In front of these tectonic lines lie the extremely ancient fragments of 
table-land which surround the Tellow sea ; in front of these again, lie ihe 
feland arcs. We may now attempt to compare these arcs more closely with 
the mainland. 

The origin of the liu-Eiu arc cannot be discussed until the structure 
of Formosa is known in greater detail. 

The arc of south Japan is connected, in Naumann's opinion, with the 
northern part of the rias coast of south China. This arc curves past 
the fragments of table-land surrounding the Tellow sea, and is joined in 
a peculiar syntaxis by the arc of north Japan. 

Next, however, so far as we can judge at present, the arc of the middle 
part of Yezo, which we may suppose to represent a part of the arc of 
Sakhalin, arises outside the arc of north Japan, so that the latter is 
ini^rted, so to speak, between the coast chain and the arc in front of it 
Still further towards the exterior lies the arc of the Euriles together 
with east Kamchatka ; towards the interior, on the other hand, the fragment 
of west Kamchatka. 

We cannot say that these arcs are syntactic in a deep re-entrant angle, 
like the Asiatic chains on the Jehlam, or like the Armorican and Variscan 
mountains between Douai and Valenciennes, or like the Ural and the arc 
of Nova Zembla on the Konstantinov-Kamen. The only exception is the 
sharp syntaxis in the middle of Honshiu ; here the bend is really similar 
to that of the above examples, yet there is a difference owing to the 

^ G. von Helmersen in A. T. von Midd^ndorTs Reise in den ftusseisten Norden und 
Oflten Sibirieng, I, 4to, St. Petenbnrg^ 1848, p. 219, Atlas, pi. zvii. 
' E. R^duz, Nouvelle G^ograpbie univeraelle, VI, 8vo, Paris, 1861, p. 860. 
' Ricbihofen, FtOirer fUr Forscbungsreisende^ p. 605. 



f^^r-WTT^'*^. " T|«r5a 



CH.in] BLOCKS BETWEEN THE FOLDS 195 

approach of the Shichi*to chain, and the north Japanese arc reproduces so 
exactly the arc of south Japan in its main feature, that we are reminded 
of the conditions on the stnut of Ormuzd where a secondary notching of 
a similar kind appears to be present (I, p. 426). 

The island arcs lie against and behind each other at an obtuse angle, 
and this is the more striking since the succeeding arc of the Aleutian isles 
shows a far greater independence, and as we proceed in the direction of 
America we recognize a much greater contrast to the arcs, and a greater 
resemblance with normal syntaxis. 

Apart from the Philippines, the relations of which are still unexplained, 
these arcs are seen to stand in very intimate connexion with the arcs of 
the Asiatic continent. Asia consists of an obstructive fragment of Indo- 
Africa — the peninsula of India, which does not now concern us, and of 
a great piece of the earth's crust folded to the south. The folds, however, 
are interrupted and separated by platforms which lie between them like 
rigid blocks, although in the platforms themselves we may also recognize 
the traces of much older folding in the same direction. 

We shall show later, with greater detail than we have yet done, that 
the Him&laya actually terminates on the Brahmaputra. There are chains 
lying behind the Him&laya joining the meridional chains of Yunnan which 
pass the end of the Himalaya and are continued in the Malay arc. This 
arc we have traced through the Banda islands as far as the coast of New 
Guinea. But although to the south it passes considerably beyond the 
equator, yet in a tectonic sense it lies wholly behind the Himalaya, or 
if we were to number the great folded ranges from the exterior inwards, 
the Him&laya would receive the number 1 and the Malay arc the number 2, 
The encounter of the two ranges above the wedge-shaped fragment of 
Shillong is thus different from the syntaxis on the Jehlam. 

We have already observed that our insufficient knowledge of Formosa 
prevents us from forming an opinion as to the arc of Liu-Eliu, but the 
south Japanese arc probably issues from the folds of south China, exactly 
as the Malay arc does from the chains of Tunnan, and it lies lehind the 
Malay arc in the same sense as this lies behind or within the Him&laya. 

In the same sense the chains of east Siberia lie behind or within those 
of south China, and, notwithstanding the peculiar insertion of the north 
end of the north Japanese arc in Tezo mentioned above^ we must regard 
all those arcs which stand in relation to the east Siberian chains as lying 
further towards the interior. With regard to this point the absence of 
a more detailed knowledge of Sakhalin and of central Tezo is much to 
be r^retted, for the map shows a much more exact correspondence of 
the coast range Sikhota-Alin with the distant arc of the Kuriles than 
with the arc of Sakhalin lying between them. 

Thus the east Asiatic coast does not resemble a series of independent 

% 



y 



r 



196 THE BORDEKS OP THE PACIFIC OCEAN [pabtiii 

ranges advancing towards the sea, but rather a stupendous vvrgation 
extending over the wlwle breadth of Euradaj the successive divergence of 
the same folded systems which, closely crowded together in the interior 
of the continents, form the great and lofty highlands. In this divergence 
each of the great branches shows near its extremity, i.e. towards the 
Ocean, a tendency to recurve to the north, and thus arise the island festoons 
of east Asia. 

13. The arc of the Aleutian idande. In scarcely any other part of the 
Pacific outline is the tendency towards the arc-like disposition so sharply 
expressed as in the zone of volcanos and mountain fragments which pro- 
ceeds from the Commander's islands through the Aleutian islands, the 
peninsula of Alaska, Eadiak and Eenal', and separates Behring sea from 
the Ocean. Orewingk, in 1850, published a geological description and map 
of this region, based on the observations of his time, which may still 
be consulted with advantage; he compared it to a knotted rope hung 
between the rocky pillars of America and Asia, which, sinking under its 
own weight, had bent its supports towards one another^. Grewingk 
represented the contrast between the Aleutian and the American directions 
of strike by means of trend-lines, and DalFs subsequent investigations 
appear to indicate that between the Mackenzie and the upper Yukon, in 
about lat. 64"^ N., the Aleutian and the west American chains meet in syn- 
taxis at a fairly acute angle ^. The chain running to the north-east which 
reaches this part of the syntaxis is called by Grewingk the Tschigmitge- 
birge ; Dall terms it the Alaskan range. It lies on the interior side of the 
principal range, which extends from the north coast of Cook's inlet through 
Alaska and the island arc. Cook's inlet itself and the strait of Shelechov 
lie in the strike of the mountains, somewhat like the gulf of Ancud and 
the channel of Moraleda in the south of Chili. The gulf of Tshagatska 
(Prince William sound) and the mouth of the Copper river mark the place 
where the course of the coast is determined by the re-entrant angle of 
syntaxis. Whether the Bomanzov mountains in the north between fort 
Yukon and the Arctic Ocean, or the upturned Palaeozoic beds which 
Beechey observed near cape Lisbum, must be regarded as northern ranges 
parallel to the Aleutian arc can hardly be determined at present 

^ C. Grewingk, Beitrftge zur Kenntniss der orographisclien and geognostischeh 
Beschaffenbeit der NW.-Etiste Amerikas mit den anliegenden Inseln, in Yerii. M in. Ges., 
St. Petersburg, 1850, 8vo, p. 215; E. von Eicbwald, Geognostiscb-pal&ontologiache 
Bemerkungen dber die Halbinsel Mangiscblak und die AlSutiachen Inseln, 8vo, St 
Petersburg, 1871. 

* W. H. Dall, Alaska and its Resources, 8vo, Boston, 1870, p. 286 ; cf. furtber, Amer. 
Joum., 1868, 2nd ser., XLV, pp. 96^99, and 1876, 3rd ser., XI, p. 242 ; C. A, White, On 
a small Collection of Mesozoic Fossils collected in Alaska by Mr. Ball, Bull. U.S. GeoL 
Surv. Territories, 1884, No. 4, pp. 98-103 ; for the Yukon, also Krause, Zeitscbr. Erdk., 
Berlin, 1883, XVIII, p. 348 ; Schwatka, Joum. Amer. Geogr. Soc, 1884, XVI, p. 374 
et passim. 



CH. in] THE ALEUTIAN ISLANDS 

The breaking up of the mouDtain chain which is accomplished 
isolated advance of the peninsula of Alaska, and then by the divi 
the prolongation of the peninsula into the groups of islands, is 
example of a phenomenon which we have observed from Kamchi 
the Eoriles, from Arakaa to the Nicobars, and on other coasta 
the Pacific structure. Here, however, the orographic form of the pe 
is modified by a form of erosion peculiar to the north. This is repr 
bj extremely deep transverse valleys, which ia Norway are called 
in AlmTfft ■ perenofisi,' i. e. carrying places or ' portages ' as they an 
in Canada, thoogh this last term is used in a more extended a 
include the low watersheds of the abraded Archaean shield. They 
same deep valleys as are described by Boas on Cumberland peninsula 
Uiey connect the corresponding fjords of the two coasts (II, p. 32). , 
just like Cumberland peninsula, is cut across many times by the 
farrows. The first runs from the most northerly part of tiie pe 
through the great lake of Uiamna, and Qrewingk enumerates fiv 
perenossi ; lakes frequently lie in their course, and Dall was informi 
some of them are so low that in the whole transit it is scarcely ne 
to lift the boat out of the water. 

Archaean rocks are known on the south-east coast of Eenal and 1 
as well as at several localities in Alaska ; at Unalashka and in » 
the western islands as far aa Attn, metamorphic rocks and ancie 
phyry are mentioned as forming the foundation. In the Comm 
islands the foundation is Archaean. At cape Nunakhalkak, on the 
east coast of Alaska (about lat. 58° 20' N.), Finart found fossils 
Trias ^ ; on the east coast Jurassic deposits are known, which, as it a 
belong to difierent stages of the middle and upper Jurassic, and the 
beds appear at various points on the east coast as well as net 
Holler on the north-west coast of the peninsula. 

The Tertiary formation of this region deserves particular attenti 
renewed investigation. In many places, particularly in Cook's in 
Tertiary leaf-bearing beds with lignite occur which distinguish the 
r^ons in eo peculiar a manner. Heer has described their flora '. 
extend as far as cape Tolstoi in Morton sound. Above these leaf- 
beds Dall observed, near Nulato, on the lower Yukon, a brown sai 
with Crepidula, Ostraea, and other marine mollusca of extinct 
This sandstone has a fairly wide extension. On the Shumagin 
which lie off the south-east coast of Alaska, DaU found resting on 
or granite, first very much altered quartzite, then blue sandy ahali 
beds of lignite, silicified wood, beds of conglomerate and leaves of Fl 

* P. PiacIieT, Snr qnelqnes ftwailea de I'Alaaka, in A. L. Finart, Voyages k 
Notd-Onert da TAm^qae, 4to, 187&, I, p. 82. 
■ 0. He«r, Flora Foaila Alaikana, 4to, Stockholm, 1869 (in Syeusk. Akad. Hoi 



198 THE BORDEKS OF THE PACIFIC OCEAN [PABrm 

then conglomerates and sand with Sequoia, and above this the broivn 
marine sandstone with Crepidula, vertebrae of whales, oysters and wood 
perforated by borings. The Tertiary beds are traversed by basalts \ 

These deposits, which appear to be of middle Tertiary age, are so far 
the only trace of a connexion with the marine Tertiary beds which have 
been mentioned as occurring in Spitzbergen and the east of Greenland 
(II, pp. 69, 73). They must be carefully distinguished from other marine 
beds which contain living species or a fauna very closely allied to the 
present, and which are regarded by many observers as also of Tertiary age, 
These occur at many points of the Aleutian arc, as well as in Kamchatka, 
Sakhalin, and on the Pribylov islands, resting horizontally against the 
older rocks of the coast, and Grewingk has justly emphasized their 
striking resemblance to the shell-bearing beds of Beauport near Quebec in 
Canada. 

In conclusion, a zone of mighty volcanos takes part in the structure of 
the Aleutian arc. It would be superfluous to enumerate them. With the 
giant Iliamna and the Ujakushatsh (or Burnt mountain) on the west side 
of Cook's inlet, this volcanic zone makes its entrance into the continent 
of America. It is very active; several great eruptions occurred in the 
course of the eighteenth century: during one of them, in 1796, the new 
volcano. Saint Johann Bogoslow, was formed, which rises from the sea west 
of the northern point of Unalashka ; finally in December, 1883, the mouth 
of Cook's inlet was the scene of violent volcanic phenomena. 

14. The west coast of America, The lofty mountain ranges which 
border the Pacific coast of America have been already described. The 
contrast between their direction and that of the Aleutian islands is very 
striking. In Chatham straits, east of Sitka, Blake observed mica-schist in 
an almost vertical position striking parallel to the coasts and from the 
works of Canadian geologists we have learnt that Vancouver and the 
Queen Charlotte islands must be regarded merely as the outer chains of 
those Cordilleras which traverse the north-western part of America. 
Further to the south we reach the great lava fiood of Washington and 
Oregon, beneath which only isolated fragments of the downthrown Cor- 
dillera are visible. It is in the Cascade range that these eruptive products 
attain their greatest height; and the overwhelmed forests beneath the 
lava recall the similar occurrences of Tertiary age which are so widely 
distributed in high latitudes. This volcanic region extends as far as mount 
Shasta, and some observers include in it the range of Lassen's peak which 
joins the sierra Nevada. 

We here approach the region of the Basin ranges, distinguished by its 

^ Ball, Note on Alaska Tertiary Deposits ; Amer. Journ., 1882 j 8rd ser., XXIY, p. 67. 
' T. A. Blake, Topographical and Geological Features of the NW. Coast of America ; 
Amer. Journ., 1868, 2nd ser., XLV, pp. 242-247. 



CH. Ill] WEST COAST OF AMERICA 199 

peculiar structare. More or leas meridional chains were formed by folding, 
then followed faults, which run almost with the strike of the folds, and 
along these the folded country has been let down. Diller has shown that 
the structure of the Basin ranges extends to the northern part of the 
sierra Nevada. This, too, is cut through between Sacramento and Honey 
lake by two faults running almost with the strike, which form American 
valley and Indian valley with a downthrow on the east, the dip being 
directed mainly to the west. On the west side of Honey lake lies a third 
line of fracture, which corresponds to the east border of the sierra. The 
outermost part of the west border of the mountains appears to be over- 
turned \ 

The difference in structure between the Coast ranges and the sierra 
Nevada, indicated by earlier investigations, is shown by fresh observations 
to be less than was supposed. In the north, Diller has found fossils of the 
Carboniferous limestone west of Shasta valley in the region of the Coast 
ranges; and, further south. White has shown the correspondence of the 
Aucella beds of the Coast ranges with those of the sierra. Becker regards 
both mountain ranges as forming part of the system of western cordilleraa 
Since the lower Cretaceous, and very probably since a far earlier period^ 
the whole region between the Wahsatch and the Pacific coast has, according 
to Becker, been the scene of the ' recurrent, if not constant action of lateral 
compression, exercised practically in one and the same direction '.' 

The distinction between the movement of folding and the formation of 
great fractures along which collapse takes place is becoming recognized 
with ever-increasing clearness. In the north, between the Siskiyou moun- 
tains and the Cascade range (south Oregon), the Cretaceous formation, 
according to Whitney, is upturned. On the western side of the sierra 
Nevada the upper Cretaceous lies horizontally in the foothills. In the 
Coast ranges the middle Tertiary beds have also been involved in the fold- 
ing. The age of the folds, as far as the position of the beds gives evidence 
of it, thus differs locally. The faults and subsidences, it is true, are also 
of different age, but the way in which they intersect the folded bands 
leaves no room for doubt that they are younger than these. Most observers 
hold the view that they are very recent, and that even the movements 
taking place at the present day are a continuation of the great process of 
subsidence. Qilbert's observations on the more recent dislocations along 
the Wahsatch fault have already been mentioned (I, p. 578). It is asserted 

> J. 8. Diller, Notes on the Geology of North Galifoniia ; Ball. U.S. Geol. Sary. 
TenitorieB, No. 88, 1886, pp. 878-387. 

* G. F. Becker, Notes on the Stxatigraphj of Galifoniia, Bull. U.S. Geol. Snnr. 
Tenritories, No. 19, Washington, 1885, p. 212 ; Cretaceons Metamorpluc Rocks of California, 
Amer. Joom., 1886, 8rd ser., XXXI, pp. 848-857, &c. ; for the palaeontological evidence 
of the Aucella beds, G. A. White, On the Mesozoic and Cainozoic Palaeontology of Cali- 
BqIL n.S. Geol. Sorv. Territories, No. 15, 1885, pp. 7-82. 



200 THE BORDEES OF THE PACIFIC OCEAN [PAKrm 

that the earthquake of March 26, 1872, on the east border of the sierra 
Nevada (I, p. 74) was accompanied by a dislocation, and Beyer describes 
ice-polished granite cliffs on lake Fordyce which have been cut through by 
post-glacial faults^. Within the downthrown area of the Basin ranges, 
and particularly in the northern part of south Oregon, Russell has traced 
fresh indications of quite recent movements, not yet covered by vegetation, 
and extending for great distances along ancient fault-lines. They traverse 
recent terraces and cones of d^ris, and may attain a throw of fifty feet ; 
Russell thinks that some of them have been produced within the last few 
years, and he gives instances from Surprise valley (long. 120"* W., lat. 
41'*-42** N.), which is often visited by earthquakes ^. 

In Central America the cordillera of the Antilles is broken off towards 
the Pacific Ocean (I, p. 642). 

Through the whole of South America as far as cape Horn, the course of 
the coast, as we have already seen, is determined by the mountain chaina 

^ E. Beyer, Zwei Profile, durch die Sierra Nevada; N. Jahrb. Min., Supplement, 
IV, 1886, pp. 291-326, map. 

' I. C. RuBsell, A Geological Reconnoissance in South Oregon ; IV Report of the 
Ann. Rep. U. S. Geol. Surr., 1882>1883, 8vo, 1884, pp. 431-464, in particular pp. 442 et 
Beq. The attempts made to draw definite conclunons as to the persistence of tectonic 
movements from the behaviour of rivers do not seem to me at present to have led to 
convincing results, for the question is affected by displacements of the strand-line, which 
must be regarded as independent of these tectonic movements ; cf. J. le Conte, A Poet- 
Tertiary Elevation of the South Nevada shown by the River-beds, Amer. Joum., 1886, 
3rd ser., XXXII, pp. 167-181. 



201 



CHAPTER IV 

COMPARISON OF THE ATLANTIC AND PACIFIC 

OUTLINES 

The Atlantic stractare. The Pacific structure. Their dlBsimilarity. BiBtribution of 
islands and volcanoe. Oyerthmsting of the depressions. Progressive completion of the 
Mesozoic series towards the Pacific coast. 

The oontants of the two preceding chapters have now placed us in 
a position to appreciate more precisely the distinction which exists between 
the Atlantic and Pacific ooastfi. 

Let na first recall the strange approach to a symmetrical arrangement 
which manifests itself in the Atlantic region. 

In the north stands the wedge-like mass of Qreenland ; on each side of 
it lies an arm of the sea. On the west the mainland begins with an 
imposing range of very ancient gneiss, which borders the west coast of 
BaflSn bay and Davis strait, and extends nearly as fitr as the strait of Belle 
Isle. On the east a similar chain of gneiss forms the greater part of the 
islands and coasts of northern Norway ; it appears to find its continuation 
in the gneiss of the outer Hebrides. 

Within the Hebrides follow the pre-Devonian overthrust folds of the 
Caledonian chain; I know of nothing similar in America. 

We now encounter in the west the Canadian shield with its fiat 
bedding, and the shallow pan of Hudson bay surrounded by a girdle of 
glint lakes. In the east we have the Baltic shield, with the Palaeozoic 
beds lying equally fiat upon abraded Archaean folds, and the shallow 
Baltic pan likewise surrounded by a girdle of glint lakea It is only by 
regarding certain fiat-bedded patches as representing the border of the 
Canadian shield, that it can be said to reach the open Ocean at all, and 
then it forms only a very short stretch of coast along the strait of Belle 
Ifile. The Baltic shield is nowhere in contact with the Atlantic Ocean, 
properly so called. On the other hand both shields reach the Arctic 
Ocean; Coronation gulf in the west, the Varanger fjord in the east, 
and to a certain extent also the gulf of Onega, present the characters 
of glint bays. 

Beyond the strait of Belle Isle follow the jagged rias coasts of New- 
foundland, New Brunswick^ and Nova Scotia, where the folds of the 
mountain ranges which were elevated towards the end of the Carboniferous 
period sink beneath the sea. These mountain ranges are folded to the 



202 ATLANTIC AND PACIFIC COMPAEED [PABrm 

north-west and north, and their outer border coincides with the right bank 
of the lower St Lawrence and the east shore of Belle Isle strait. The 
gulf of St. Lawrence forms the boundary of these folded ranges; the 
island of Anticosti does not belong to them, but to the border of the shield. 
To this rias coast corresponds in Europe the rias coast of the Armoricaii 
range, which is also in its main features post-Carboniferous, certainly pre- 
Permian in age, and which extends from the Shannon nearly to La 
Bochelle. This range is cut through by St George's channel and the 
English channel ; it is folded to the north-east, north, and finally to the 
north-north-west 

In Europe the next part of the Atlantic coast further south is occupied 
by the last outposts of the Pyrenees ; we may regard the formation of this 
chain, together with the Alps, as subsequent for the most part to the 
subsidence of the pre-Permian ranges, the course of both being to a large 
extent determined by the trend of those blder mountains and by the 
position of their faulted margins. 

The overthrusting of the northern border appears to be less marked 
than in other European ranges ; towards the west, in the north of Spain, 
this border reaches the sea, but it has collapsed in the Basque provinces 
(I, p. 289 ; n^ p. 121). Then comes, along the north-west coast of Spain, 
the fractuied margin of that group of structures, almost unique in their 
arrangement, which form the basin of Asturias, and indeed in such a 
manner that the several members of this basin become more and more 
perpendicular to the coast as they strike out to sea. This basin also is 
of pre-Permian age, and it looks as if it were about to repeat itself in the 
sudden bend of the Betic cordillera in the straits of Gibraltar. 

It is the fragmentation and reconstruction of the folded ranges which 
have given to the coast of western Europe its variety of form. These 
processes did not occur in North America, and there a mediterranean sea 
bounded towards the Ocean by the bjtc of the Antilles lies much further 
south than in Europe. In Europe the arc of the Antilles finds its precise 
homologue in the arc of the Betic cordillera. 

The Alleghanies and all the other folded ranges of eastern North 
America as far as Newfoundland are folded away from the Atlantic Ocean 
and turn their inner side and their oldest rocks to it. Our increasing 
knowledge of the mountains of Brazil shows us more and more clearly that 
the folding in all the chains of this region is also turned away from the 
Atlantic in the direction of the Andes, so that the innermost zones lie next 
this Ocean. Consequently the arc strikes out to sea in cape Corrientes and 
Staten island. 

In west Africa the end of the great Atlas is unknown ; the south shows 
fractured table-land. 

With the exception of the cordUlera of the AtUHUs and ihe mountain 



CH.IV] ATLANTIC CHARACTEES 203 

t 

fragment of Gibraltar, which form respectively the boundary of the two 
mediterranean seas, the ovier side of a folded range nowhere determines the 
ouUine of the Atlantic Ocean. The older folded ranges which extend from 
Maine to Newfoundland turn, it is true, their outer side to the lower 
St. Lawrence and the strait of Belle Isle, but where they reach the great 
Ocean they disappear beneath it. The inn^r sides of folded ranges, jagged 
rias coasts which indicate the subsidence of Tnountain chains, fra^ctured 
margiTis of korsts, and fractured table-land form the diversified boundary 
of the Atlantic Ocean, 

The same structure also characterizes the coast of the Indian Ocean as 
far to the east as the mouths of the Ganges, where the outer border of 
the Eurasian chains meets the sea. The Erythraean trough, the fracture 
of the Quathlamba in Natal, as well as that of the Sahy&dri in India, and 
the structure of Madagascar, the faults of which have been recently 
described by Cortese, indicate that the structure of this region has been 
determined by tabular fractures ^. It is only in the Persian gulf that some 
of the outer Iranian zones reach the sea. 

The west coast of Australia likewise exhibits Atlantic structure. From 
the arrangement of the folded ranges we know that they turn their convex 
side to the Pacific Ocean, and a comparison with South America shows that 
from the continent of Australia as far as New Zealand and probably New 
Caledonia, a more or less concentric system of folds is present, turned 
towards the Pacific side. 

The west coast of Australia thus presents much the same relations as 
that of the east coast of Brazil. 

The borders of the Pacific Ocean may be divided for convenience into 
five parts. 

The first is formed by the arc of the Aleutian idands. While the 
north of the Atlantic Ocean is characterized by the Archaean mass of 
Greenland, here on the other hand is swung an island arc, affording with 
its folded Mesozoic beds and its inner zone of active volcanos a complete 
contrast of structure. The syntactic angles of the arcs in the north Pacific 
Ocean at once recall the relations of the mountain chains of India 
a, p. 462). 

The second part consists of the west coast of North America, from the 
golf of Kenai down to the coast of Mexico. The Queen Charlotte islands 
ate regarded by Canadian geologists as an outer chain of the Cordilleras. 

The third part is South America) it begins in Guatemala, where the 
Cordillera of the Antilles strikes across Central America, is divided into 
two by a syntaxis in the gulf of Arica and prolonged in an arc to the 
south beyond cape Horn. This part is characterized by the coast Cordillera, 

^ £. Cortese, Una escunione al Madagascar, Boll. R. Com. Geol. d'ltalia, 1887, XYIII, 
pp. 129-134; and OsBeryauoni geognosiiche sul Madagascar, torn, cit., pp. 181-191. 



^ 



204 ATLANTIC AND PACIFIC COMPARED [PARrm 

where the stratified series does not appear to begin till the Neocomian ; it 
recalls in many respects the coast chains of California, as well as the 
Nicobars and Andamans, which have a similar structure. 

The fourth part is formed by the arcs of eastern Asixi, which we have 
recognized in many of their most important segments as the recurved 
extremities of the great chains of central Asia; they are not syntactic 
after the manner of independent arcs, but lie one behind the other, forming 
the termination of a series of chains folded in the same direction and 
belonging to a single system. The great Malay arc marks the southern 
border of this system. Timor and Soemba lie outside it ; whether these 
two islands should be included in the Australian region must for the 
present remain undecided. 

As the fifth part we regard the Australian chains together with New 
Zealand and New Caledonia. An opportunity has already presented itself 
of comparing the Australian chains of the Flinders and Adelaide ranges as 
far as the east coast with the meridional sierras in the west of Cordova. 
The absence of middle Tertiary deposits over the whole of the eastern 
coast of AustraUa and Van Diemen's Land, in contrast to their rich 
development on the south coast of Australia and in Bass strait, leads to 
the supposition that the continent east of the existing coast has only in 
comparatively recent times subsided to the great depths which are now 
known to occur in this region. 

To complete the boundary of the Pacific Ocean we shoidd here add, 
perhaps, a south part, the Antarctic^ but the accounts at our disposal of 
this region are too fragmentary to be of any value. The syntaxis of an 
arc lying outside the Australian system with the south side of New 
Zealand (II, p. 183, Fig. 16) is not without suggestiveness. In the more 
southern latitudes, H. Beiter has concluded &om the observations of 
travellers, that the regions corresponding to the Pacific Ocean and those 
lying further to the east, from the Balleny islands and Victoria land to 
the south Orkneys, are constructed on the Pacific type ; the other r^ons, 
however, on the Atlantic type ^, It is possible that this conjecture may be 
confirmed some day by fresh discoveries. 

With the exception of a part of the coast of central America in 
Guatemala^ where the bending Cordillera of the Antilles has siu/nk in, the 
whole border of the Pacific Ocean, wherever it is known in any detail, w 
formed of mountain chains folded towards the Ocean in such a manrier 
thai their outer folds either form the boundary of the mmidand iiseif or 
lie in front of it as peninsuUis and island chains. 

No folded range turns its inner side to the Pacific; no table-land 
reaches the shores of this Ocean. 

^ H. Reiter, Die Sddpolarfrage und ihre Bedentuog fQr die genetische Gliederung der 
Erdoberflftche, 8to, Weimar, 1886, 84 pp., map. 



CH,iv] ATLANTIC AND PACIFIC ISLANDS 205 

In the AUaTUic region the shallow bajsin of Hudson bay on the 
Canadian shield corresponds to that of the Baltic and the golf of Bothnia 
on the Scandinavian shield The irregular arms of the sea around the gulf 
of St. Lawrence, between the rias coasts of the ancient mountain folds of 
New Brunswick, cape Breton, and Newfoundland, occupy a position similar 
to that of the English channel which crosses the Armorican folds. The 
great inbreak of the foreland of the Antilles occupied by the gulf of 
Mexico finds no counterpart in the European region, where a dominant 
character is the repeated formation of mountain chains; the Caribbean 
sea represents the Mediterranean, unless, indeed, we prefer to compare it 
with the basin of Asturias. 

The volcanic islands stand in groups, like the Azores, the Canary 
islands, and the Cape de Verde islands, or run in straight lines as in the 
prolongation of the Cameroons in the gulf of Guinea. The coral reefs of 
the I^uscadives and Maladives, as far as the Chagos islands, also lie in 
a straight line. Yolcanos arranged in curved lines only occur in one 
of the two regions in which the Pacific type encroaches on the Atlantic 
region, namely in the Antilles. 

Kerguden, according to the investigations of Studer, Moseley, and 
Benard, consists of successive sheets of lava, chiefly basic K These rise in 
great terraces one above the other ; the trachytes and phonolites appear to 
be older than the basalts. Between the basic flows there exist in some 
places intercalated beds of lignite, sometimes with trunks of great conifers, 
probably the relics of ancient forests. These facts show that Kerguelen is 
the remains of a once more extensive land and that the outflows did not 
take place beneath the sea. This resemblance with the Faeroe islands and 
Iceland is very remarkable. 

In the Pacific region, on the other hand, and particularly in that part 
of it which belongs to Eurasia, the arc-like disposition and the 'hinder 
seas' within the arcs are characteristic. We may doubtless regard that 
branch of the Mediterranean which lies to the north of Candia and Cyprus 
as a representative of these hinder seas of eastern Asia ; the gulf of Pegu 
is another example of this relation which recurs under various forms as far 
as Behring sea. It is less marked on the west coast of America. The 
boundaries of the strait lying on the inner side of Queen Charlotte islands 
are particularly incomplete. Tlte gulf of California, on the prolongation 
of the valleys of San Joaquin and the Sacramento, as well as the gulf of 
Corcovado on the prolongation of the longitudinal valley of Chili — both of 
them clearly determined by the strike of the mountain chain — difler to 

^ T. Studer, Geologische Beobachtungen auf Eerguelenland, Zeitschr. deutsch. geol. 
Gee., 1878, XXX, pp. 327-B50, map, pi. xt ; H. N. Moseley, Notes of a Naturalist on the 
* Challenger,' pp. 184>215 ; A. F. Renard, Notice sur \a g^ologie de Tile de Kerguelen, 
Boll. Mu8. B. d*HiBt. nat de Bruzelles, 1886, IV, pp. 228-272, pi. ▼. 



206 ATLANTIC AND PACIFIC COMPAEED [PABrm 

some extent from the hinder seas of eastern Asia. They do not lie behind 
the Cordilleras, but between the coast chain and the principal cordillera, or 
entirely within the coast chain, as happens in the south, and these pecu- 
liarities are repeated with slight modification in the strait of Shechelov 
and Kenal sound. 

The mountain arcs which characterize the boundaries of Eurasia, and in 
particular of eastern Asia, exhibit the arcuate plan in almost every possible 
stage of existence. Beginning with a completed cordillera with volcanos 
alined on its inner side, as in Italy — or upon its inner side and in the midst 
of the cordillera itself, as in north and south Japan — we pass to a cor- 
dillera cut off by the sea and prolonged by a mere chain of islands, with 
volcanos rising out of the sea, as in the chain of Pegu (coast of Arakan, 
Nicobars, and Andamans) with the volcanos of Puppa Doung, Chouk Talon, 
Barren island, and Narkondam, on the inner side; and then to an arc 
having no visible attachment to the mainland, with fragments of the 
cordillera forming an outer chain of islands and volcanic mountains an 
inner chain, as in the liu-Kiu; next to a volcanic arc with the merest 
remnants of a cordillera, as in Java ; and finally to a volcanic arc alone, as 
in the Kuriles, unless, indeed, the east of Kamchatka represents a part of 
the cordillera of these islands. 

The case is similar as regards the outer Australian arcs. If we accept 
Dana's views as to the trend-lines of the Polynesian islands we must regard 
not only New Caledonia with the Lousiade group, but also the Loyalty 
islands and the whole zone of the New Hebrides, the Solomon islands. New 
Ireland, and the Admiralty group as outer arcs of the Australian region. 

Whatever may be the true relation of this chain of islands to the 
mountain ranges of Australia, it is certain that in advancing towards the 
Ocean the arrangement of the islands becomes less and less arcuate* To 
the east, on the west coast of central America occurs, as we have seen, the 
only case in which a folded range is cut directly across by the Pacific 
coast. This is the cordillera of the Antilles, with its well-known re- 
semblance to the Cordilleras of South America. Precisely at that point 
where the arcuate continuation of this chain might be expected to meet 
the principal chains of South America, lie the volcanic Galapagos — a 
sporadic group presenting the same form of association as occurs elsewhere 
in the Atlantic region (I, p. 143). 

We have quoted Bichthofen's observation that 'certain facts point to 
step faults on the grandest scale as forming the boundary of the Pacific 
basin ^.' Drasche believed that it would be sound geology to draw the 

^ F. von Richthofen, Fahrer f&r Forschungsreiseiide, 8vo, Berlin, 1886, p. 605. We 
may remark in passing on the concordance in the results obtained by inyestigators in 
describing the continental margins ; see C. V^lain, Les Volcans, 8yo, Paris, 1884, p. 126, 
figs. 42 and 43, and J. Walther, tJber den Bau der Flezuren an den Grensen der 



CH. IV] WATEESHED OF THE GLOBE 207 

western boundary of the Pacific Ocean outside the island arcs from 
Ejimchatka through Japan and onwards from New Zealand through 
Auckland and Macquarie islands to Victoria land in the Antarctic region^. 
In any case we see how the volcanic circle of the Pacific resolves itself into 
a series of arc& Many of these lie inside the fragmentary remains of 
a Cordillera ; others, especially ihe giant volcanos of South America, stand 
upon the cordilleras themselves. Further, we can now clearly recognize 
that the hypothesis of the formation of folded chains by the thrusting 
from the Ocean towards the land of the border of a sinking basin or 
geoBjmdinal is in no way founded on fact. There is no geosyndinal in 
the world greater than the Pacific, and the mountain ranges, so far from 
bemg turned away from it, are turned in the opposite direction, facing it, 
and BO afford an example on the grandest scale of the general tendency to 
oyerthrust the depressions (I, p. 148). 

The march of the folded ranges on the borders of the Pacific is 
•ooompanied both in North and South America by a general descent of 
the mainland in the opposite direction, that is towards the Atlantic, and 
on both continents almost the whole of the rainfall flows away on this side. 
The same is true, although not in so marked a manner, of the other 
continents. A. von Tillo has described the line which runs from cape 
Horn along the west coast of America to Behring strait, and from there 
through the middle of Asia across the coast of Syria east of the Nile to 
the cape of Good Hope, as 'the chief watershed of the globe V When we 
indude Arabia and India also in the region bounded by coast built on the 
Atlantic type we realize how disproportionately large is the volume of 
fresh water which it supplies to the sea. 

A last important character should here find mention. This is tJie 
completion of the marine Meeozoic series which we perceive as we approach 
tke Pacific coasts. On the Atlantic coasts nothing similar is to be observed. 
This fact, so important in the history of the Oceans, will clearly appear in 
the following chapters. 

Eontinente, Zeitschr. Naturwiss., Nene Folge, Jena, 1886, XIII, Taf. zii, fig. 8. Walther 
lays particular Btrees on Scrope*B view, that eruptions predominate in those parts where 
the convexity tarns downwards. 

^ R. yon Drasche, Uber pal&ozoische Schichten anf Eamtschatka nnd LuEon ; N. Jahrb. 
Min., 1879, pp. 265-269. 

' A. Ton Tillo, Ein Wort fiber die Hanptwasserscheide der Erde ; Peterm. Mitth., 1887, 
HXin, p. 101. 



208 



CHAPTER V 
PALAEOZOIC SEAS 

Introduction: The abyssal region; cycles; thickness of the sediments. The 
North Atlantic continent. Upper limit of the Silurian. Universality of the middle 
Devonian transgression. The Carboniferous system. Paralic beds. Alternation of coal 
beds and marine sediments. Transgression of the Carboniferous limestone. The 
Permian system. Gondwdna land. General Survey. 

The object of the next three chapters is not to give a complete survey 
of the results of stratigraphical geology. From the earliest times the 
starting-point of all the various theories which relate to changes of land 
or sea has been the remarkable fact that we find in the midst of the land 
the remains of shells or fishes belonging to the sea. This was the problem 
which first pressed for solution ; now another question confronts us, and 
we must seek to determine, by means of the facts ascertained in the study 
of strata, whether the conditions under which inundation and emei^nce 
took place in early times were of such a nature as to compel us to seek 
their explanation in secular oscillations of the continents, or whether, on 
the contrary, this hypothesis is incompetent to explain them. We are 
thus led to consider the successive changes in the extension of the seas, and 
at the same time the various nature of the sediments. 

While dealing with the first point we shall have occasion to refer 
in greater detail to the fact mentioned at the conclusion of the last chapter 
— that the Mesozoic series grows more complete as we proceed from the 
interior of the contin^its towards the Pacific Ocean. In the second place, 
in treating of the Carboniferous period, the formation of the Coal-measures 
must be carefully discussed, for, in explanation of this, it has been assumed 
without hesitation that the solid land may have oscillated up and down 
as many as a dozen times, or even more. In our treatment of the Mesozoic 
aera and subsequent periods we must pay particular attention to the nature 
of the calcareous formations. 

The treatment of the subject will consequently be very unequal, and 
only such facts will be chosen as have an important bearing on the question 
at issue. 

A study of the history of the seas must certainly be based on a 
knowledge of their existing state. This knowledge, however, is very 
fragmentary. The expeditions undertaken in recent times to explore the 
deep regions of the ocean have brought to light so many new and 



CH.V] NATUEE OF THE SEDIMENTS 209 

unexpected facte, that it is evident we have as yet scarcely crossed the 
threshold of great discoveries in this field of investigation. But these 
explorations are attended by every kind of difficulty ; isolated soundings, 
separated by great intervals, are only obtained here and there. The slopes 
of our mountains, on the other hand, reveal to the eye, in innumerable 
exposures, vast masses of sediments accumulated at various depths in the 
andent seas, together with the remains of the animals which peopled them. 

Among these ancient formations we may distinguish those which have 
been deposited by evaporation from solution, such as gypsum and rock- 
salt; next clastic sediments, i. e. those which have been mechanically 
transported and then deposited on the bottom, such as sand and mud ; and 
finally organic formations, such as coral reefs, which have been constructed 
by living creatures. In addition to these three chief groups, two others 
must also be mentioned. They both resemble clastic formations in many 
important respects, but their material is not derived from the solid land, 
which, as is generally admitted, is the source of clastic deposits, but from 
the depths of the sea itself. These are, first, volcanic ashes and scoriae, 
which rise from the bosom of the earth and spread themselves out over 
the floor of the ocean; and secondly, the remains of the hard parts of 
calcareous or siliceous organisms, such as Radiolaria and Qlobigerinae, 
which, together with fragments of shells and other d^ris, form a silt of 
organic origin, and these have contributed largely to the sediments of 
ancient seas, now forming a visible part of the earth's crust. 

The chemical composition of beds of rock-salt and its associated 
minerals, shows that the substances held in solution by the sea were the 
same in the past as in the present. But as soon as we leave these products 
of evaporation, formed in lagoons or in more or less enclosed arms of the sea, 
and turn to the depths of the Ocean, the difficulties of comparison increase. 

In existing seas the temperature remains constant from a certain depth, 
which diminishes towards the poles, down to the bottom. The sun does 
not penetrate into this abyssal region, the manifold influence of diurnal 
and seasonal changes or difierences in latitude does not affect it. The same 
climate prevails in the depths of all the oceans, and a uniform fauna 
characterizes them, exposed to no change of conditions except differences 
of pressure, which correspond to difierences of depth. Thus it appears, as 
recorded by Wyville Thomson, that the' CJiaUenger in dredging below 
500 to 600 fathoms everywhere brought to the surface a fauna identical 
in all essential features, in the Indian Ocean as in the Southern Ocean 
down to the Antarctic circle ; in the Pacific Ocean, both north and south ; 
and in the different regions of the Atlantic Ocean ^. 

^ C. Wyville Thomson, Report on the Scientific Resalts of the Voyage of H.M.S. 
CJuittenger, during the years 1878-1876, under the command of Captain G. S. Nares and 
Captain F. T. Thomson ; Zoology, I, London, 1880, pp. 48-50. 

SOLLAI II P 



210 PALAEOZOIC SEAS [pakt in 

It is not until we reach the shallower zones of the sea that the 
differences in warmth and illumination begin to manifest themselves^ the 
distinction of climates makes itself felt, and zoological provinces assume 
definite limits. The diversity in external conditions of existence increases 
as we approach the shore, and with it the multiplicity of living forms 
themselves ; and beyond the margin of the tides, on the dry land, where 
the rays of the sun have only to penetrate the atmosphere, and where lungs 
have taken the place of gills, not only the variety but the variability 
of external conditions reaches its maximum. We can now more easily 
understand the extraordinarily wide distribution of certain marine faunas 
in early times; it is also comprehensible that on dry land and in the 
shallower zones of the sea the displacement or even the extinction of 
faunas may be brought about by altered conditions of existence. But this 
renders it all the more difficult to explain the fact that the cosmopolitan 
abyssal fauna, exposed to no difference of temperature and hardly even sub- 
ject to change of place, should yet have likewise suffered' transformationa 

Thus the whole of the animal world which forms the biosphere falls 
into two chief groups, according to its habitat, of which one is under the 
influence of the sun, the other withdrawn from this influence. The first 
group comprises the inhabitants of the land, of fresh water, and of isolated 
zones of the sea ; the second group is the abyssal circle of living things. 

The absence of sunlight in any region inhabited by Uving beings, 
whether in a cave, or the depths of a lake, or the bottom of a well, or 
finally in the abysses of the Ocean, betrays its effects, in the most diverse 
classes of the animal kingdom, most obviously by modification or atrophy 
of the eye. A great number of animals in the abyssal regions are blind, 
so were many Trilobites of the lower Silurian, and particularly of the 
Cambrian seas — a correspondence which has been repeatedly pointed out; 
the phenomenon, however, is so peculiar that it deserves to be illustrated 
by a few examples. 

A little blind Arthropod, Gecidotaea stygia, which lives in the caves of 
North America, is closely related to AeeUus coTnrrkunis, which possesses 
sight and does not inhabit caves ; it looks like an emaciated, badly armed 
Asellus which has lost its eyes. Its introduction into the caves must be 
later than the time when the river terraces were being formed in the 
valleys of the United States. But it has been shown by Packard that 
the optic nerve, together with the ganglia, is completely absent in all the 
specimens of Cecidotaea which he has examined ; nevertheless, in some in- 
dividuals a small dark spot is visible on the exterior, which under the micro- 
scope is seen to represent the extremely rudimentary remains of the eye. In 
the majority of individuals, however, there is no trace of such a structure \ 

^ A. 8. Packard, On the Stracture of the Brain of the Sessile-eyed Crustacea ; Mem. 
Nai Acad. Sci. Washington, 1885, III, pp. 99-110, pi. 



K 



CH.V] EYES OF DEEP-SEA ANIMAIiS 211 

Niptargns, a blind geans of the Qammaridae, occurs in the cavee of 
Carinthia, as well aa in the underground waters of many springe ; and 
Forel has foond it in the depths of the lakes of Geneva and Keach&tel. 
Hambert has expressed the opinion that the species living in the Swiss 
l^ea have not been produced by a modification of species possessing sight, 
sach aa Oavvmante pulex, but 
owe their origin to the immi- 
gration of forms inhabiting 
Buhterranean water '. 

Nnmerona other facts hav- 
ing an important bea^ng on 
the origin of isolated faunas 
in lakes might be cited in proof 
of such a migration of small 
animals living in subterranean 
waters; and it is an open 
quertion whether the American 
Ceddotaea etygia might not 
have been an inhabitant of snb- 
terraaean water before it occu- 
pied caves, and even before the 
fatmation of the river terraces. 

The New Zealand ■ f rost- 
fiah,' Lepidatus coiufofua, which 
has very lai^ eyes, has been 
described by von Lendenfeld 
as a deep -sea species. The 
migration of an inhabitant of 
the shallower zone into the tm- 
iUamined depths is a very alow 

process, requiring for its ac- Fw. is. (V»Kwma N^ftama, Qu«r.-H«n»v. («nor 

complidiment, ««.rdmg to ''!;S,XSS"2u„u. o„„ .. »p. St Vln«.t, 
Lendenfeld, many successive ut. S6*47' H., long. 8° 38' w., in i,090 fsthomi. 
generations. If the eye were 

origumlly well developed, it might gradually increase in size ; but if, on the 
contrary, weak or unaUe to adapt itself, owing to over-rapid change of 
balntat, atrophy and blindness might result *. 

We sometimes find, not only towards the upper limit of the abyssal 

t A. Humbert, Deccriptioa da Niphtuyua pufMnus, nx. Fordii, BnU. Sac. Vand. Sci. 
DaL, 1876, XIV, pp. 2T8-S98 and pL; abo Ai«h. Sci. phjs. nat. Gen^re, 1877, noav. »it., 
vol. I, lOa, pp. 58-75. 

* R. TOO Lendenfeld, Note on the E^ea of Beep-aea Fiahea ; Proc. Linn. Soc N.S.W., 
Sydney, 1885, IX, pp. 609, 700. 



212 PALAEOZOIC SEAS . [part ni 

region but much below it, even down to great depths, some w nip i ri n 
poBsessiug well-developed eyes with others completely blind, and also 
even, here and there, some which show clearly that the eye has become 

enlarged. Cyatiaoma 
NeptuTms (Fig. 18), 
with eyes which meet 
in the middle line, 
lives at great depths 
in the Atlantic, yet 
has several times been 
captured at the sur- 
face during the night ; 
it appears to belong 
to that lai^ group of 
Pro. 19. IValopUMalymu armigtr. Will. Suhm (after Wille- pelagic animals which 
moM-Suhm). i, xi _e 

Magnified tvro diUDetera ; between cape Palmaa and Saint- COme tO tue SUrtace at 
Paul,iD2,B00fathoin8; irOnauticalmilnaeaataf Saint-Paulin niirht uid minim hv 
1,600 Aithoma ; alao in ut SC *1' 8. long., 20" 66' W., about 100 ~* ana reinmi oy 
nautical milea weet of Triatan d'Acunha group, in 100 fathoma. day tO the depths. 

The SchiEopod Petal- 
ophthtUmus armiger, described by Willemoes-Sulim, is still more character- 
istic (Fig. 19). Its eye-atalks terminate in vesicular enlargements, which 
are formed of chitin and show no 
trace of the structure proper to 
sight. The eye has been destroyed 
by its enlargement '. 

Polycheles and Willemoeeia are 

closely alUed to the Jurassic genus, 

Eryon; the two living forms are 

inhabitants of the deep sea, and 

Willemoesia has been found both in 

the Atlantic and Pacific Oceans at 

a depth of 1,900 fathoms. The eyes 

in both these genera are so small 

and well cxmcealed that for a long 

(.S?^S,.„'SS„T*^ ™'- '"^ ti™ "-ey w.™ overlooked, b.t 

Kagnified four diameten ; 460 fathomi ; Spence Bate affirms that in early 

m rero, ea d ea. stages of development the embryo 

of Willemoesia possess organs of sight constructed ou the usual Crustacean 

plan* (Fig. 20). 

We will now turn our attention to the observations on the eyes of 

' R. yon WiUemoes-Sabm, On some Atl&ntic Crustacea from the 'Challenger' 
Expedition ; Tnni. Linn. Soc. London, 1S75, 2nd ser„ I, pp. 23-59 ; pi. 
< 0. Spence Bate in Challenger, Kanative I, 2nd purt, p. 524. 



r 



CH, V] 



BLIND TRILOBITES 



213 



Trilobites made by Bamode, long before onr present knowledge of the 
deep-eea faona had thrown light on their special significance. 

In 1873 Barrande summed up all that was then known on the eyes of 
Trilolntce aad anived at the following results ' : — 

Among the Tnlobitee of Bohemia there are six genera of which all the 
speciee are blind, and six which inclnde both blind species and those pro- 
vided with eyes. The earliest or ' primordial ' fauna contains, out of 
27 species, seven without eyes ; the second fauna, 25 species out of 127 ; 
and the third fauna only one species out 
of 206. The blind species appear in the 
argiUaceons sediments, more particularly 
in the Cambrian slates, and in the argil- 
laoeoos deposits d,, d^, and d^, of the lower 
Silurian D, hut not in the beds which are 
predominantly qnartzose, and the single 
blind species of the upper Silurian is 
found in ai^iUaceous limestone. There 
ate, however, forms with quite abnor- 
maUy large eyes, snob as Aeglina and 
Remopleurides, and it is just theee large- 
eyed species which are found associated 
with the completely eyeless ones and 
once lived along with them '. In the 
ease of Aeglina armata, as in that of 
the English species Aeglina mirabilie, the 
enlargement of the eyes is so far advanced 
that they meet from side to side in the 
middle, and Barrande calls this the cydo- 
p«in form (Fig. 22). 

It is obvious that the eye of Trilo- 
bites in the Silurian period was exposed 
by disuse to an atrophy analogous to that experienced in other classes 
of animals in later times, and that here too the result was blindness, 
either by direct atrophy or by enlargement, which is as it were an attempt 
of the eye to maintain its existence. Acidaspia mira (Fig. 23), which 
occurs in the upper Silurian limestone, shows a stalked eye like that of 
existing Crustacea, well furnished with oi^ans of sight ; the eye of Aeglina 
(Fig. 22) may be compared with that of Cysiisoma Neptunus (Fig. 18), 

' J. Banande, STst^me Hilnrien du centre de la BohGine, Sappl^ment an vol. I, 4to, 
Piagae. 1S72, pp. 155-164 and 195-197. 

* 'Poaitait-on penser que lee yenz de cee TrllobiteB ^taient devtin^i k aoppl^ pu 
lenn dimentiona extraordinairea i la jaiblene de la lumifece tmiamiM k tiaveii lea 
(MIX tronblei ? ' Banande, torn, cit., p. 162. 



Fk.21. OmooipiMi'Ca SulMri, Sohloth. 
(kftar Burande). 

Primordial stage ; Oetuetz, Bo- 
bemik ; blind. 



214 PALAEOZOIC SEAS [part m 

which well deserves Burande's designation ' Cyclopean ' ; it exhihits that 
process of enlai^ment which is exemplified by the ' &oet-fiah ' of New 
Zealand, and with the development 
of peduncles would lead to the blind 
goggle eye of Petalophtkalmvs ar- 
miger. Oonocepkalitea Sulzeri,6iiBilj, 
which is quite blind and occurs in 
the Cambrian slatee of Bohemia as- 
sociated with ConocephxdUea tiriatua, 
a species provided with eyes, corre- 
sponds with many examples of com- 
plete absence of any external vestige 
of an eye which ore known in the 
deep sea, in deep inland waters, in 
subterranean water, and in caves. 
These species which have become 
' blind lived amidst otheia which were 

Fts. 2-i. Cjolopean 'Ejta (after Burande). provided with eyCS, aS is also the 

Toaek ; e, 4. At^faa arviata, d, , neftr LeUkow. the deep sea and at the bottom of the 

lake of Geneva, as has been observed 

by Forel '. Barrande has also found that in the youngest individuals 

of Trinudeus BwMandi, a blind species of the lower Silurian, there is 

a little wart in place of the eye 

in the middle of the cheek, which 

disappears with age ; we are thus 

reminded of Willemoesia, the 

Lemaides and other an i mala 

which possess in their youth the 

^. rudiment of on eye, destined 

jS with the advance of age to be- 

S come partly or entirely lost In 

the blind species Trinucleua 

omatvs a young form possessing 

eyes has not so far been observed 

(Tig. 24). 

Fia. as. Aeidagpit miro, Barr. (arter Bamuide). The numerical statements of 

^^Pp.rSU«m.,I,L«l.iHB.J..„l.. SU11«1 g^^j, „ to the distributioi. 
of blind Trilobites in the several 
stages may possibly be very much modified by later discoveries. Barrande 
himself observes that the preponderance of blind species in the Cambrian 

' J. Forel, Fanne profonde dn lac L^man ; Verb, •chweu. nat. Qes. sa Chur, 187S-1874, 
p. 186. 



i 



CB.V] BLIND TKILOBITES 

deposits of Bohemia is to be attributed to Agnoa 
divergent genns. On the other hand, it maet not be 
other Cambrian genera, snch as Hydtocephalos, Pi 
England Anopolenos ', certainly poseees a facial 
elongated ocular tubercles, but bo far as I know fac 
observed, except in a very few species, if indeed in i 
Silurian the veiy large eye, also elongated and actuc 
genus Bemopleoridea indicates, even by its size, men 
blindnee& Indeed so far Conocepkalitee etricUvs appt 
E{>ecie8 of the ' prhnordial ' fauna of Bohemia in which 
to observe the facets. Whatever light further investij 
this question, so much is already evident, that even 
among Silurian animals the twofold process which 
leads to blindness was already in active operation. 

From this we may conclude that the cause of 
the blindness was the same as among the existing 
inhabitants of the deep sea, namely the absence of 
Bonlight ; at the same time, however, it appears 
that the (ddeat known fauna of the Cambrian 
^stem most be r^arded as a traiiaformed fauna, 
and thus presupposes the existence of a still older 
fauna of which we have no knowledge. Just as the 
ancient salt depoaita testify to a correspondence 
between the past and present, both as to the nature 
of the substances contained in solution and the order 
in which they were extricated by ciystallization, so 
eyes in the oldest known organisms gives us reason 
penetration of stmligbt into the once-existing seas took 
conditJons and exercised on the organs of vision the i 
the present day. 

So far we may pursue the comparison between the ] 
with some degree of certainty; but as soon as we at 
with precision the limits of t^e abyssal region and to e: 
difficulties present themselves. Observations on the d 
penetrates, made by tink'Pg a white disc, cannot be aa 
unce the perceptive faculty of these strangely modified 
to us, and many of the animals which have become blin 
tkalniMa armiger mentioned above, have the habit of c< 
of the sea at night. Thus it happens that the expn 

' H. Hicks, Note tm the Qeniu Anopotentu; Quart. Jonm. 
pp. 477-482, pi. Blind anitnKls which present no exterior tm 
poweM a cODcealed mdiment, have been long known a* oc( 
Cmttacea, 8vo, London, 1858, Introduction, p. zxxi et paasim. 






216 PALAEOZOIC SEAS [part in 

/ understood in most different senses. T. Fuchs, who has given particular 

attention to this question and who by his exact knowledge of the Tertiary 
deposits is especially qualified to form an opinion, lays great stress on the 
influence of light, but places the limit between the littoral zone and the deep 
sea at a depth of only 40-50 fathoms. Other authorities differ greatly ; 
GUnther, for instance, puts the approximate limit of the deep-sea fauna at 
a depth of 500 to 600 fathoms \ 

For those studies of a general nature in which the expression is 
employed, it seems to me most convenient to draw the line where the 
greatest difference of the most general kind makes its appearance, and that 
is where the diversity of climates ends and the universal fauna b^ins. 
This fauna is not subject to those laws of distribution which dominate the 
shallower zones, and changes in external conditions must affect it in an 
altogether different manner. 

The shallower regions may now be further divided into several zones, 
and they are distinguished also by the great variety of their sediments. 
But the undoubted fact must not be overlooked that terrestrial plants 
and clastic sediments may be carried away into great depths ^ The 
presence of beds of little pebbles of variegated quartz in the Cambrian 
slates of Ginetz, which contain the blind Trilobites mentioned above, is 
a striking example of this fact. Indeed it must be admitted that in many 
cases the belt of clastic materials which surrounds many parts of the 
continents descends into the region of the abyssal fauna, and that it did so 
even in the ancient seas. 

This clastic belt, however, is neither equally broad nor of equal thick- 
ness,' and where it recedes or thins out the pelagic sediments introduce 
themselves with continually increasing purity. As a rule they are 
calcareous, though under certain circumstances limestone may also be 
deposited at trifling depths or even at the sea-level, for the conditions 
which govern the formation of marine limestone are very varied. There 
are massive coral reefs in which living organisms not only secrete the 
carbonate of lime, but also determine the form of the reef as a whole. 
These, however, constitute but a small fraction of the calcareous deposits, 
which for the most part are regularly stratified. Limestone may idso be 
/ formed by organic debris, i.e. by the breaking up under the action of the 

waves of the hard calcareous parts of molluscs, corals, and other marine 
animals ; this disintegration may proceed so far that after a storm a broad 
band of milky water surrounds the limestone reef, as Alexander Agassiz 

^ T. Fuchs in numerous papers ; in particular in Verli. k. k. geol. Reichs., 1882, 
^ pp. 55-68 ; Sitz. zool.-bot. Gas. Wien, April 5, 1882; and by the same, Welche Ablage- 

rongen haben wir als Tiefseebildungen zu betrachten ? N. Jahrb., 1882, 2. Beilage-Band, 
b pp. 487-584. 

' Examples in Fuchs, Welche Ablagerungen, &c*, p. 498 et seq. 



CH. v] OSCILLATION 217 



in the case of the keys of Florida. Finally limestone may be 
formed by the showering dovm on the bottom of the sea of delicate 
calcareous shells, sach as the tests of Globigerina ; but at extremely great 
depths carbonic acid occurs in such quantity that these delicate shells are 
dissolved and no limestone is formed. There are various signs by which 
limestone beds reveal the action of the atmosphere and a temporary 
emergence from the sea. One of these is the interspersion of fragments of 
light red clay or of argillaceous limestone in the white limestone, such as 
occurs for instance in the Bermudas. Of a similar nature are the residues 
of red earth left on the solution of limestone, whichi are met with on so 
many coral islands, and in the dolinas of the E^rst, where they form the 
'terra rossa.' Another indication is the presence under certain circum- 
stances of phosphates, the origin of which may be similar to that of the 
existing beds of guano. These few examples may su£Sce to show how 
various are the phenomena. But they do not exhaust the difficulties, and 
there is still another question for solution, since some beds of carbonate of 
lime contain a small, and others a considerable, admixture of carbonate of 
magnesia, and beds of dolomite may alternate with beds of limestone. 

Thus the most careful observation is necessary in the study of stratified 
limestones. 

Many years ago Murchison expressed the opinion that the. middle stages 
of every geological formation consist of limestone, and this idea was worked 
out in greater detail by Hull in 1862 for several formations, particularly 
the Carboniferous. The predominance of the terrigenous, that is clastic, 
elements carried down from the land should, according to Hull, lead us to 
recognize phases of oscillation of the land, and ke thus distinguishes three 
stages: — 

Upper stage . . . movement . . . detrital formation. 

Middle stage . . . repose calcareous „ 

Lower stage . . . movement . . . detrital „ 

At the same time he concluded from the distribution of the clastic 
materials in EIngland that we may assume land to have existed to the west 
and north-west during a series of geological periods ^. 

The simpler stratigraphy of the United States has led American 
geologists to adopt similar views. Newberry, in 1860, starting from the 
Cretaceous transgression of the west, arrived at the conception of ' cycles of ' 
deposition,' i.e. a periodic return of similar conditions of deposition, or in 
other words an alternation of deposition in shallow water and in the open 

^ E. Hull, On Iso-diametrio lines as a means of representing the Distribution of 
sedimentaiy Clay and Sandy Strata^ as distingnished from Calcareous Strata, &c.; 
Quart. Jonm. GeoL Soc., 1862, XVIII, pp. 127-146, pi. 



218 PALAEOZOIC SEAS [pact ra 

sea. In 1874 this theory was applied in detail to the Palaeozoic sediments 
(I, p. 13). 

Newberry distinguished the following elements in such a cycle: — 
(1) coast, (2) region outside the coast, (8) open sea, (4) retreating sea; or 
sediment 1, mechanically transported; 2, mixed; 3, organic; 4, mixed. 
Within a larger cycle smaller cycles owing to oscillation would arise, as is 
supposed to have happened in the Carboniferous, and strictly speaking only 
three elements can be distinguished in a cycle, namely, two deposits in 
shallow water separated by deposits in the open sea. The Palaeozoic aera 
would include four great cycles \ 

Other investigators in America have followed in the satne direction, 
most prominently Dawson in 1868 ^. 

A study of the Tertiary series of Belgium had led Andr6 Dumont to 
recognize that the intercalation of thick beds of coarse pebbles indicates 
some particular change in physical conditions ; he therefore attempted to 
determine the limits of the Tertiary stages by means of these pebble beds. 

Rutot and van den Broeck have renewed this attempt They endea- 
voured to show that in secular, continental subsidence (positive movement), 
the margin of the sea as it advances towards the interior covers the land 
with a layer of coarse beach pebbles {gravier d'iw/merewn) ; and on this, as 
the movement continues, finer sand, and finally the clay ^ the deeper marine 
zones, is laid down ; if the oscillation then passes into its opposite phase, 
i.e. if negative movement succeeds, the clay will again be followed by sand, 
and this by an upper layer of rolled pebbles (gravier cCimereian), which it 
is true is often not fully developed or is destroyed by erosion. This series — 
pebbles of submergence, sand, clay, sand, and pebbles of emergence — is called 
a complete ' cycle s^imentaire ' and regarded as equivalent to a complete 
secular oscillation. The clay thus takes the form of a lens thinning out 
towards the land. In this particular case no calcareous formation of any 
importance results ; only clastic materials are deposited. The cycles may 
be incomplete or the corresponding deposits removed in part by erosion. 
The author has cited as an example the lower Tongrian marine stage, con- 
sisting of Tg la (pebbles), Tg 16 (sand), Tg Ic (clay), Tg Id (sand), which all 
belong to the zone of Ostrea ventilabrum; then comes Tg 2a (sand with 
Cytherea semistriata), Tg 26 (clay with Cytherea iificrassata), Tg 2c (sand 
with Cerithium pLicatum) \ 

^ J. S. Newbeny, Circles of Deposition in American Sedimentary Rocks ; Proc. Amer. 
Assoc. Adv. Sci. Portland, Maine, 1873, 8yo, Salem, 1874, pp. 185-196. 

' J. W. Dawson, Acadian Geology, 2nd ed., 1868, pp. 135-138. 

' A Rutot, Les ph^nom^nes de la sedimentation marine ^tudi^s dans leurs rapports 
avec la stratigraphie r^gionale. Bull. Mus. K Hist nat. Braxelles, 1883, II, pp. 41-88 ; 
E. van den Broeck, Note sur un nonvel mode de classification et de notation graphiqne 
des depots g^ologiques bas^ sur T^tude des ph^nomtoes de la sedimentation marine, torn, 
cit., pp. 341-369. 



CH. v] CYCLES AND THEORY OF ELEVATION 219 

CoTMndcrmtions of this kind lead to the supposition of very frequent and 
Yeiy regular alterations of the ooast-line which are difficult to reconcile 
with the principles of the elevation theory, and Newberry chooses expres- 
sions such as : advance of the sea and retreat of the sea, or the same words 
as were employed by Brogniart, Omalius d'Halloy, and others before the 
prevalence of the elevation theory (11^ p. 18). 

According to these views the limits of a system, sometimes even of its 
subordinate stages, might be regarded as an expression of the negative 
phase, while the positive phase would correspond approximately to the 
middle of the formation. But it is a fact that the nomenclature created in 
Europe, and with it the principal limits of the stratigraphical systems, have 
been found applicable in. all geographical latitudes and in the most distant 
regions of the globe. It is impossible, however, that a cycle of this kind 
can affect in the same sense the whole surface of our planet. We must 
suppose that positive areas in one part are opposed by negative areas in 
another. The theory of elevation, in particular, is inconsistent with the 
conception of a positive or negative movement extending simultaneously 
over the whole earth, or even over a very large part of it ; diversity of the 
movement is characteristic of this theory. 

The thickness of the sediments themselves, however, sometimes amounts 
to thousands of feet. In every individual case, these sediments, as they 
increased in thickness at any given place where a section has been studied, 
must necessarily have produced a constant diminution in the depth of 
the sea. 

If the level of the coast-line remain wholly unchanged they must 
finally reach the surface of the sea, and affording the observer evidence of 
shallow water and life under littoral conditions, may lead him to the 
mistaken inference of a negative movement. Every negative movement is 
strengthened in its effects by the progress of sedimentation, and every 
positive movement increases the depth of the sea only by so much as it 
exceeds the thickness of sediment forming at the same time and place. The 
same considerations apply to an oscillatory movement. It may then happen 
that the thickness of the sediment, limestone for example, may equal the 
positive excess, or Sed. = Pos. — Neg. The upper surface of the sediment 
will then correspond to the highest level of the positive phase, that is to the 
sign e or 971 in the table on p. 25 ; and it is in no small degree probable that 
this surface will lie above the sea-level and lead us to infer an elevation of 
the land. We shall meet with this case in recent limestone formations and 
coral islands. 

The process may be represented somewhat more exactly. The thickness 
is often so great, that by increase of the local attraction it raises the sea- 
level a perceptible amount, a. At the same time over the whole earth 
sedim^it is being continually carried down into the sea; thus the sea is 



220 PALAEOZOIC SEAS [PABTin 

gradually forced to overflow its shores, and for any given interval of time 
by a quantity <r. The formula which represents the increase in the depth 
of the sea is therefore 

(Pos. + a + <r) > (Neg. + Sed.) 

and for a diminution in the depth of the sea : 

(Pos. + a + <r) < (Neg. + Sed.). 

In this connexion I cannot repress a doubt that in some of the most 
frequently repeated statements as to the measured thickness of clastic 
deposits, particularly of the Carboniferous system, some exaggeration in 
the figures is to be feared. It is not always permissible in fact to take the 
sum of the several thicknesses, as for instance in limestone formations, for 
this method assumes that the sediments in question were deposited verti- 
cally, one above the other, while some at least were really formed on talus 
slopes advancing towards the sea, and thus, to a certain extent, beside each 
other or resting against one another. In this way we might obtain figures 
for the thicknesses of clastic sediments which would far exceed the truth ; 
still in any case these thicknesses, even if strictly estimated, would certainly 
amount to many thousands of feet. 

1. The North AUaTUic continent, A belt of clastic sediments affords 
a basis of hypothesis as to the position of the continent from which it has 
been derived. In North America we meet with the greatest thicknesses of 
Cambrian and Silurian shale and sandstone in the vicinity of the Atlantic 
coast, namely in the east and south-east of Canada, in the adjacent north- 
eastern part of the United States, and in the Appalachians. Thence the 
thickness diminisheis towards the Mississippi ; shale and sandstone diminish 
in importance, while limestone increasingly predominates. 

We see something similar, although not with equal distinctness, in the 
north of Europe. In the British Isles great thicknesses of sandstone and 
shale are present, and the same is the case in the mountains of Norway ; in 
Sweden and the Baltic provinces the thickness rapidly diminishes, and 
in the Silurian the limestone predominates more and more. 

This fact has led American geologists to conjecture that the continent 
which furnished these sediments occupied the site of the existing North 
Atlantic Ocean. In the same way in England, Oodwin-Austen, and after 
him Geikie, regarded the crystalline rocks of Scandinavia, with their 
continuation into the Hebrides, as the remains of the Silurian continent, 
and Hull has defended with great zeal the theory of a vanished Palaeozoic 
Atlantis \ From the following, however, it will appear that this continent 
persisted as such up to a very recent epoch in the history of the earth. 

^ E. Hull, On the Geological Age of the North Atlantic Ocean ; Trans. R. DnUin Soc 
1885, new ser., Ill, pp. 805-820. 



CH.V] THE NORTH ATLANTIC CONTINENT 221 

When we attempt to study in greater detail the manifold subdivisions 
of the marine beds as they succeed each other in the east and west of this 
primaeval continent, we are at once struck by the far more fortunate 
position of the observer in America as compared with his colleague in 
Europe. In the west, during the accumulation of these deposits, orogenetic 
movements took place only in the Oreen mountains, and flat-bedded Palaeo- 
noic sediments still extend over vast areas far away up to Kansas and 
Nebraska. Even where folding followed later, as in the east of Canada 
and the AUeghanies, we have as a rule no difficulty in correlating the 
stratified succession with that of the more distant parts of North America. 
In Europe, on the other hand, as soon as we have crossed the narrow zone 
of the Hebrides we encounter the Caledonian folds with their great over- 
thrusts, and upon the ruins of these worn-down Silurian mountains the Old 
Bed sandstone Ues horizontally. If we proceed further into the interior 
of this continent we again find the ancient marine deposits still for the most 
part folded and fractured ; it requires, therefore, a combination of persever- 
ance, exactitude and sagacity, and these in no small degree, to establish 
a correspondence between the members of the stratified series in regions 
at all remote from one another. It is not till we reach Sweden and the 
Russian plain that we meet with stratigraphical relations as favourable to 
study as in the United States. 

Let us again turn our attention to that region. 

The thickness of the Palaeozoic sediments decreases as we proceed from 
the Atlantic Ocean towards the Mississippi. In the Rocky mountains also 
it is trifling, but beyond them, towards the Wahsatch and Uinta, the 
sediments thicken out again very rapidly, as they also do in the eastern 
part of the Basin ranges, according to Clarence King up to about long. 
117^ 15^ W., and then they disappear. In California only the Carboniferous 
limestone is known. Beyond long. 117'' 15' W. there thus lay in all pro- 
bability a second continent, which bounded the Ocean on the west \ 

As early as 1859, James Hall had described in a masterly survey the 
dasflifieation and distribution of the Palaeozoic beds in the east and centre 
of the United States. The years which have since elapsed have confirmed 
his results in all essential points, and they form the basis of the comparisons 
which follow here ^. The investigations of Canadian geologists and those 
carried out in the great ranges of the west complete the picture. 

In Newfoundland, New Brunswick, and at Braintree, Massachusetts, 
there are beds, chiefly shales, in which are found species of the genus 
Conooephalites, both blind and provided with eyes, as well as a number 
of other Trilobites, which closely agree in general character with the 

» C. King, U. S. Geol. Expl. 40tti Par., I, pp. 127-248, map. 

' J. HaU, Geological Survey of New Tork; Palaeontology, III, 4to, Albany, 1859, 
pp. 1>96. 



222 PALAEOZOIC SEAS [pam m 

* primordial ' fauna of Ginetz and Skrey in Bohemia ^ In America also 
they represent the oldest animals so far known. This group of strata is 
called the St. John's group. In the whole central part of the United 
States they appear to be absent, and it is only when we reach the Wahsatch 
and Eureka, in Nevada, that we again encounter them ^ 

In the east, from Labrador and Newfoundland as far as New York and 
Vermont, and again in the west, in Utah and Nevada and in British 
Columbia on the east border of the Bocky mountains, this group is 
followed, according to Walcott, by a middle member of the primordial or 
Cambrian series, the Georgia group, with an independent fauna. limestone 
plays a leading part in its composition ; this group again is not yet known 
in the central part of the United States \ 

The distribution of the last member, the Potsdami aandsUme, is much 
more extensive, and may be traced from Belle Isle through Canada down 
into the north-east of the United States, through Pennsylvania and Virginia 
and further to the south-west into Tennessee, where it becomes thicker and 
alternates with dolomitic limestone; then in flat-lying beds still further 
into Texas, and again to the north in an elongated zone which lies hori- 
zontally on the south border of the Canadian shield from lake Huron 
through Wisconsin and Iowa *. From Iowa, where, according to D. D. Owen, 
several subdivisions may be distinguished, the zone of sandstone continues 
to the west. It crops out in the Black hills from the plain of Dakota, and 
then as an almost continuous border, sometimes gently inclined, sometimes 
steeply upturned or even inverted, surrounds the Archaean ridges of tiie 
Bocky mountains. Further west in Nevada the fossils of the Potedam 
group lie high above those of the earlier stages. 

The stratified series of Eureka begins according to Hague and Walcott 
with about 1,500 feet of brownish -white quartzite, the upper part 
of which becomes shaly and calcareous; the first traces of the middle 
Cambrian or Georgian fauna appear here. There then follows about 
8,000 feet of grey limestone, containing the same fauna, to which, at the 
summit, species of the upper Cambrian or Potsdam fauna are added. The 

^ G. F. Mathew, Illustrations of the Fauna of the St. John's Group, Trans. R. Soe. 
Canada, Montreal (in vols. I, II, and III, 1884-1886) ; G. D. Walcott, On the Cambrian 
Faunas of North America, Bull. U. S. Geol. Sunr. Territories, 1884, No. 10, p. 49, pL 

* A. Hague, Abstract of Report on the Geology of the Eureka District, Ann. Bep. 
U. S. Geol. Surv. Ill, 1881-1882. by Powell, Bir., 8^0, Washington, 1888, p. 248 et aeq. 
C. D. Walcott (Palaeontology of the Eureka District ; U. 8. Geol. Surv. Monograph VIII, 
1884) has described this fauna and in particular the abnormal development of Olendlu$ 
HawUli. 

' Walcott, Second Contribution to the Studies on the Cambrian Faunas of N. America ; 
BuU. U. S. Geol. Surv. Territories, 1886, No. 80, pp. 729-952. 

^ J. Hall, Note in Proc. Amer. Assoc. Adv. Sci., XXXI Meeting, held at Montreal, 
8vo, Salem, 1883, pp. 63-65. 



CH. v] THE PKIMOKDIAL COAST 223 

limestone is followed by 1,600 feet of shale, 1,200 feet of limestone, and 
300 feet of shale. All these belong to the Potsdam stage. 

A veiy singular picture is presented to us by this ancient Ocean. 
Shore deposits, with cracks produced by the drying action of the sun and 
filled with millions of valves of Lingula, form the series known as the 
Potsdam sandstone, which still lies flat on the south border of the Archaean 
shield of Canada ; in Nevada, far to the west, these deposits swell out to 
a great thickness, while limestone and shale take the place of sandstone. 

Let us now turn to the Black mountains as described by Newton and 
Jenney. The Potsdam* sandstone, 200 to 300 feet thick, presenting at its 
base a coarse bed of beach pebbles with auriferous sand, includes some 
glauconite beds and a little limestone; it rests in discordance on the up- 
turned edges of the Archaean schist. Here we have before us an Archaean 
reef of the Cambrian sea, which in the last stages of its existence was 
completely covered with sand. Over the Potsdam sandstone, however, 
follows — not the succeeding systems of the lower Silurian, upper Silurian, 
and Devonian, but the marine deposits of tHe Carboniferous which lie 
immediately on the sandstone and apparently in complete conformity 
with it *. 

In the Bocky mountains the Potsdam sandstone is similarly overlain by 
these same deposits of the Carboniferous system. 

Ejight degrees of latitude further south and also further to the west is 
the Grand cafion of Colorado, excavated in horizontal beds of Carboniferous 
limestone. Powell and Dutton, however, state that the Carboniferous 
series, 4,000 to 4,500 feet thick, rests on the edges of inclined beds, probably 
Cambrian, which are cut off by the plane of transgression. This is an 
example, exposed for miles along the cafion, of a plain of erosion produced 
by the advance of the coast-line ^. 

Throughout the whole of Arizona also, the Carboniferous rests directly 
on Archaean or possibly Cambrian rocks, and up to the present it is only in 
Nevada that the western development as represented by the lower Silurian 
has been recognized. 

Thus the Carboniferous transgression circumscribes the region in which 
it is possible to study the post-Cambrian deposits. 

The oldest f ossilif erous sediments of North America enable us therefore 
to recognize the following : — 

The St. John's group and the Qeorgia group, the two lower members, 
are only known in the east and in the west. A wide interval separates 
these two provinces. The third member, the Potsdam group, covers them 

^ H. Newton and W. P. Jenney, Report on the Geology and Resources of the Black 
HiUs of Dakota ; U. S. Geogr. and Geol. Surv. of the Rocky Mountain Region, 4to, Wash- 
ington, 1880, pp. 80^106 and 109. 

' C. E. Datton, Tertiary History of the Grand Guion District, p. 178 et seq. 



224 PALAEOZOIC SEAS [part m 

both and extends over the whole region. In Nevada it increases in thick- 
ness and limestone makes its appearance, but in Dakota and over the whole 
of the north border, where it lies on the Archaean rocks of Canada, it is 
without doubt a coast formation deposited in transgression. This example 
shows that a strictly littoral formation, deposited in transgression, must be 
regarded not as the sign of a negative phase, but, on the contrary, as a proof 
of the advance of the sea on the land. 

Notwithstanding the close attention which has been given to the 
classification of the Cambrian beds in England, and the extreme precision 
with which linnarsson and Brogger have distinguished the several members 
of this system in Scandinavia where they are less thick, yet I dare not 
venture to make any parallel with the American beds beyond the correlation 
of the St. John's group with the deposits of Ginetz in Bohemia. A trans- 
gressive littoral group, which might be compared with the Potsdam sand* 
stone, has not been observed in Europe. Its upper limit, or the base of the 
lower Silurian, is however expressed with equal clearness on both sides of 
the Ocean by the change of fauna. 

2. The 'Upper limit of the SiXurian formation. In the east of North 
America the upper Silurian series begins with the Clinton group or the 
Medina sandstone; both show distinct signs of variable conditions and 
somewhat shallow water. Over these beds lies the Niagara limestone, 
very widely distributed and containing a rich marine fauna ; it is generally 
regarded as the equivalent* of the Wenlock and Ludlow beds in England 
and of the upper Silurian limestone E of Bohemia. This would represent 
the middle phase of the Silurian cycle. Then follow beds deposited in 
shallow water, so shallow indeed that layers of gypsum and rock-salt were 
formed over wide area& These form the Onondaga saliferous group, 
a sandy formation bearing the marks which attend the concentration of 
sea-water, yet attaining a thickness of 1,000 feet. It presents itself with 
these characters in New York State, particularly in the western part, and 
proceeds with diminished thickness, on the one hand into the south-west of 
Virginia, and on the other to Wisconsin, often broken up into patches by 
erosion. In the interior of the continent, towards the Mississippi, this 
member is characteristically absent, and the upper Silurian Niagara lime- 
stone is directly succeeded by the Devonian limestone. 

Here then we enter the open sea; but nearer to the ancient coast we 
find the saliferous group passing upwards through many alternations of 
sediment into a brownish-yellow dolomitio limestone known as the ' Water- 
lime.' This extends beyond the saliferous group into Illinois and Iowa, and 
contains in some places the remains of the great Crustacea belonging to the 
genera Eurypterus and Pterygotus. This beyond all doubt is also a shallow 
water formation. 

Let us now investigate the upper limit of the Silurian in the north of 



CH. v] SHALLOW WATER AT CLOSE OP SILURIAN 225 

Europe. In the highest parts of the Ludlow series we meet here and there 
with a *' bone-bed ' containing numerous remains of fishes drifted together ; 
then follow beds of variable character, but always littoral, which Murchison 
deeeribes as ' Passage beds ' ; these are sandstones with Lingula, or marly 
shales with fish remains and great Crustacea, notably the gigantic Pterygotua 
am^ieus ; between these again are shales with lingula and thin layers of 
sandstone. Where the earliest sediments of the Old Bed sandstone are to 
be seen superposed on the Passage beds, we observe that remains of the 
same fishes and Crustacea are continued up into them, and thus the lower 
part of the sandstone is often assigned to the Silurian ^. 

The Silurian concludes in England as in North America with an unmis- 
takable and considerable diminution in the depth of the sea. 

In the Baltic region as on the island of Oesel we find a similar condition 
of things. According to F. Schmidt, in the uppermost beds of the Silurian, 
a yellow platy dolomite, poor in fossils, makes its appearance, or a grey 
sandstone containing many upper Silurian corals and other fossils. In the 
platy dolomite Euiypterus again occurs in abundance, accompanied by 
Pterygotus, but here, as in Gothland, the beds with Eurypterus are succeeded 
by others still containing upper Silurian fossils '. 

Finally, the same succession is repeated in the river valleys of eastern 
Galicia and in the adjacent parts of Russia. F. Schmidt, who is so 
thoroughly acquainted with the Silurian deposits of the Baltic, has him- 
self visited these regions and recognized the correspondence. The upper 

^ B. J. Murchison, Siluria, 4th ed^ 1867, p. 186 ; by the same, On the discoveiy by 
Mr. B. Slimon of Fossils in the Uppermost Silurian Bocks near Lesmahago in Scotland, 
Quart. Jonm. Geol. Soc 1856, XII, pp. 15-25. For the ascent of the fossils into the Old 
Bed sandstone see B. Etheridge, Anniversary Address of the President ; Quart. Joum. 
GeoL Soc, 1881, XXXVII, p. 17a A visit to the habitat of horny-shelled brachiopods 
led me some years ago to examine a great number of the facts quoted here. Homy- 
sheiled brachiopods have subsequently been found in deep water also. The fauna of the 
deep sea has become known, and with this knowledge a complete alteration in the 
previously accepted views of the Cambrian fauna of Bohemia. At that time elevation 
and depression of the land was the accepted theory, but it was the progress of these 
particular studies which led me first to doubt the correctness of the prevailing opinion, 
next to recognize the extent of the Cenomanian transgression, and finally to results 
which were no longer in accord with the traditional method of explanation. Ueber die 
Wohnsitae der Brachiopoden ; Sitz. k. Akad. Wissensch. Wien, 1859, XXXVII, pp. 185- 
248, and 1860, XXXVIII, pp. 151-206. An account of the Passage Beds and their 
littoral characters occurs in the second part, pp. 189-191. In this memoir the 
recurrences of the lower Silurian in the United States (p. 182 et seq.), which are 
not described in detail here, are also discussed; the comparisons made there for 
the Utica shales are completely in accordance with the results of the admirable 
work of 0. D. Walcott, The Utica Slates and related Formations, Trans. Albany 
Inst, 1879, X. 

' F. Schmidt, Bevision der ostbaltischen silurischen Trilobiten nebst geognostischer 
Uebersicht des ostbaltischen Silurgebietes ; M^m. Acad. Imp. Sci. Saint-P^tersb., 1881, 
XXX, No. 1, p. 49 et seq. 

■OIXAS n Q 



226 PALAEOZOIC SEAS [part in 

Silurian is again represented in the same manner as in England and the 
Baltic provinces. Here also Eorypterus occurs in the equivalents to the 
Ludlow beds, and Silurian fossils appear to be present above the beds 
with Eurypterus*. 

We are thus in a position to follow the beds with Eurypterus from 
the centre of the United States as far as the Dniestr. In New York State 
the saliferous group underlies the platy dolomite with Eurypterus, as 
though the latter no longer corresponded with the phase of maximum 
depression of the coast-line. 

In England the Eurypterus beds pass upwards into the lowest beds 
of the Old Bed sandstone. In Gothland, Oesel, and on the Dniestr, they 
are again succeeded by a few beds with marine Silurian fossils. But 
everywhere, from Iowa to Podolia, a zone of shallow-water deposits 
distinguished by strange gigantic Crustacea makes its appearance towards 
the upper limit of the Silurian. 

In America this shallow-water zone thins out to the south-west, and the 
marine upper Silurian Umestone is directly succeeded by Devonian lime- 
stone ; in Europe it is the same, but the thinning out occurs at a greater 
distance from the original coast-line. In the best-known regions of the 
interior, that is Bohemia, the upper Silurian limestone is also succeeded 
by another limestone, which, according to recent investigations, we may 
refer to the lower Devonian. 

8. The Devonian System, In the British isles and in the Baltic 
provinces the greater part of the Devonian system is represented by the 
Old Bed sandstone, which contains fishes (among them curious, heavily 
armoured ganoids), and here and there terrestrial plants, but no coralB 
nor marine shells. This red sandstone comprises divers subdivisions of 
the Devonian; and in England, as we have seen above, its lowest beds 
may be assigned to the Silurian. 

In central Europe the case is different ; the red sandstone is absent, 
and its place is taken by slates and quartzite, also even by limestone, with 
a rich marine fauna. Our knowledge of this part of the stratified series 
has been greatly enlarged by Beyrich and Kayser's researches in the Harz. 
Eayser has shown that the fauna of the limestones which are intercalated 
with the lower Devonian ' Wieder Schiefer ' of the Harz corresponds to tiie 
upper divisions of the Bohemian series, hitherto assigned to the Silurian. 
It is on these fossiliferous localities of the Harz and Bohemia that Eayser 
has based his Hercynian stage K 



* F. Schmidt, Einige Bemerkungen Qber die podoUscli-galiziBche Silurformation imd 
deren PetrefiEikten ; Yerh. k. russ. min. Gas. St. Petenb., 1876, 2. Ser., X, pp. 1-21, mapu 

' E. Kayser, Die Fauna der Sitesten Devonablagenmgen des Hanes, Abh. z. geoL 
Specialkarte ▼. Preussen, 1878, II, Heft 4, pp. 1-293, pi. ; and by the same, Ueber die 
Grenze zwischen Silur and Devon (Hercyn) in Bobmen, Thilringen and einigen andereii 



CH. V] 



EED SANDSTONE OP THE NOBTH 



227 



Kayser justly compares the appearance of this new marine stage with 
that of the different marine faunas, which, like the fauna of the Alpine 
Trias, the Rhaetic, and the Tithonian, are imknown in northern Europe. 
They are the oceanic representatives of those periods, and their equivalents 
in the north are either sublittoral sediments or non-existent. We may 
trace the Hercynian stage, i.e. the formations of the open sea during the 
lower Devonian period, from the Pyrenees, where it is represented by 
argillaceous bedsS to the Harz^ where it occurs as limestone included in 
slates, and to Bohemia and the eastern Alps, where it is pure limestone. 
In northern Europe it is not to be f oimd. 

Having made these general observations, let us return to the Eury- 
ptems beds of northern Europe. 

It is certain that in England Eurypterus ascends into the lower beds 
of the Old Bed sandstone, and that in general the relations of these lower 
beds with the Silurian are so intimate that we must include them in that 
system. The upper part of the sandstone has been subdivided, on strati- 
graphical and f aunistic evidence, in very various ways ; but all lead to the 
conclusion that it represents a very large part of the Devonian, although 
farther to the south, in Devonshire, the Devonian already presents itself 
as a marly and in part also calcareous series, with a fairly rich marine 
fauna'. 

The red sandstone extends in many patches of flat-lying beds over 
the folded Silurian strata of the Caledonian chain into the extreme north 
of Scotland; it forms the Orkney and Shetland islands (Fig. 10, II, 
p. 78). 

In the Baltic provinces the red sandstone is likewise superposed on the 
uppermost Silurian strata; but, according to the unanimous opinion of 
Pfvgliah and Russian geologists, the fish remains of the Baltic sandstone do 
not represent the lower, but the middle and upper beds of the British 
sandstone, particularly as developed in Scotland^. Between the highest 
Silurian with Eurypterus of Oesel and the red sandstone immediately 
above it there is a gap, and the lower Devonian is not represented. 

We have seen previously that towards the north this red sandstone 
overlaps the several stages of the Silurian, so that between St. Petersburg 

G^egenden, N. Jahrb. Min., 1884, II, pp. 81-86. I here restrict the Hercynian to 
the stage F, following 0. Nov^k, Zur Eenntniss der Fauna der Etage Ff^, Sitz. k. 
bohm. Ges. Wiss., 1886, and Freeh, Zeitschr. deutsch. geol. Ges., 1886, XXXVIII, p. 917. 
The position of g^ remains for the present doubtful. 

^ C. Barxois, Sor la faune de Hont-de-Yer (Haute-Gkironne) ; Ann. Soc. GdoL du Nord, 
1886, XU, pp. 124-144, pi. 

* Geikie, On the Old Red Sandstone of Western Europe, Trans. B. Soc. Edinb., 
1878, XVIII, pp. 345-452 ; E. Hull, On a proposed Devono-Silurian Formation, Quart. 
Joura. GeoL Soo. 1882, XXXVIII, pp. 200-209, etc. 

* Geikie, torn, cii, p. 862 ; Lahusen in Schmidt, Revision, p. 80, &c. 




/ 



228 PALAEOZOIC SEAS [part in 

and Gatschina the upper Silurian is already entirely ooncealed, and only 
a narrow band of lower Silurian remains visible ; at the point where the 
Sjas enters lake Ladoga the Cambrian is alone to be seen ; further to 
the north, as far as the White sea, the sandstone rests directly on the 
Archaean (II, p. 45). In an instructive series of charts representing the 
transgressions to which Russia in Europe has been subjected, Karpinsky 
has assumed that the patches of qtiartzite which occur at Onetz and at 
Powjenez, north of lake Onega> as well as the unfossiliferous sandstone of 
the island of Eildin, near the Murman coast) and of the peninsula of 
Bubatsch, may also be assigned with considerable probability to the Old 
Bed sandstone \ Of more doubtful age are the isolated patdies of sand- 
stone in the northernmost parts of Norway; in Spitzbeigen, however, 
we again meet with the fish remains characteristic of the Old Bed sand- 
stone in the liefde bay beds, but no marine Silurian is yet known to 
underlie it (II, p. 69). In Greenland also we have noticed patches of 
red sandstone, unfossiliferous it is true, resting on the Archaean in the 
peninsula separating the fjords of Sermilik, Tunugdliarfiak, and Igalliko 

(II, p. 78). 

In the peninsula of Gasp^, however,^ and in New Brunswick, the 
Devonian fishes of the red sandstone have already been found ; here also 
they are associated with terrestrial plants, and the correspondence of botti 
plants and fishes with those of Europe is, according to Dawson, beyond 
doubt *. 

It cannot be denied that in this northern region a certain resemUanoe 
with the Indian Ocean exists, the red sandstone playing here the same 
part as the Gondw&na beds do there. 

The open sea of the Devonian period did not extend over Spitsbergen, 
nor over Gasp^, and it was probably excluded over the whole distance 
from Gasp^ to Greenland. The transgression of the middle Devonian 
sandstone over the several stages of the Silurian in northern Bussia shows, 
on the other hand, that previous to the transgression some parts of the 
Silurian had already been worn away by denudation, as had also the 
Caledonian folds. 

Let us now pass to a somewhat closer consideration of the Bussian 
transgression. 

In Livonia and Courland a broad zone of red sandstone with fish 
remains occurs, which, according to Grewingk, is about 100 meters thick ; 

^ A. Earpinskj, Skizse der physiko-geographiBchen VerhftltnisBe des europftiBcheii 
Ruflsland in den vergangenen geolog^schen Perioden ; vorgetragen in der Offenilichen 
Siizong der Akademie am 29. Dec, 1886 ; Bditr. s. Eenntn« des BnsaiMhen Beiche% 
III. Folge, St. Petersburg, 1887, p. 14, note. 

^ W. Dawion, Canadian and Scottiih Qeology ; Trans. Edinb. Geol. Soc^ 1885, Y, 
pp. 112-122 et passim. 



^.~ l^-' 



CH. v] DEVONIAN TRANSGEESSION IN RUSSIA 229 

above it on the south foUows a series of dolomite and limestone with 
marine fossils, gypsum, and beds of marl containing pseudomorphs of 
rock-salt. The thickness of this middle group is estimated by Grewingk 
at 70 meters. Upon it, forming a narrow zone, about 20 meters thick, 
lies an upper bed of red sandstone ^ 

Near EJiolm, on the Lovat, in the government of Pskov, 800 kilometers 
from the southern boundary of the Silurian zone, Earpinsky found the 
Silurian onee more cropping out from beneath this series of transgressive 
beds, and it appears that here also there is the clearest evidence of a partial 
denudation of the Silurian anterior to the transgression, and consequently 
during the lower Devonian period ^ 

Tschemyschew and Wenjukow have studied the fossils of the limestone 
and dolomite beds. They belong to the middle Devonian, and show that 
after the deposition of the lower sandstones, which are also assigned to 
the middle Devonian, a period occurred during which gypsum and a small 
quantity of rock-salt crystallized out in lagoons, but in some localities 
there, was deeper water inhabited by a genuine marine fauna. Then 
followed once more a deposition of sandstone^. 

To the east and south-east the sandstone gradually disappears ; its last 
known traces are seen in Orel beneath the middle Devonian limestone. 

The brief account by Tschemyschew shows this very clearly. The 
npper and middle Devonian with a rich marine fauna occur in the 
governments of Orel and Voronezh; far to the north their presence 
beneath the plain has been ascertained by borings. These facts have 
been represented by Earpinsky in a diagram which shows wedge-shaped 
beds of limestone thinning out into the sandstone towards the north. The 
thinning out of the limestone may be plainly seen on the Volkhov K 

To the north-east on the Uchta, which flows into the upper Petchora, 
we again encounter the middle Devonian, and above it the Domanik 
Aales. 

These have been described by Eeyserling; the name is said to be 
derived from the Russian word ' dym ' (smoke), for the quantity of bitu- 
minous matter contained in the shale is so great that it bums readily with 
a smoky flame. Petroleum exudes from the ground. Cake-shaped nodules 

^ G. Grewingk, G^logie von Liv- nod Enrland, 8vo, Borpat, 1861, pp. 9-61 ; by 
the same, Erl&aterang zur zweiten Aufigabe der geognofltischen Karte Liv-, Est- und 
Korlandfl, Arch. Nat. liv.-Esth-Eiirlands, Dorpat, 1879, ser. 1, YIII, pp. 15 et seq. 

* A. Sjupinsky, Zur Qeologie des Gk)aTemement8 Pskow ; M^l. phys. chim., 1886, 
Xn, p. 622. 

* T. TschemyBchew, Mateiialien zur Eenntniss der devoniBcben Ablagemngen in 
Bnnlaad, M4m. Com. 6^L Rnasie, 1884, I, no 8, pp. 77-81 ; P. N. We^jnkow, The 
fwnsk of the Devonian System of North-west and Central Russia (in Russian), 8to, 
St Petersburg, 1886. 

* Earpinsky, Skine, pp. 17-18, note 1. 



• ' 







280 PALAEOZOIC SEAS [pABrm 

of limestone occur in an intercalated bed of green marly day; tihey 
correspond with the upper Devonian Ooniatite beds of the Rhine dis- 
trict^. 

It is not until we reach the Ural that the lower Devonian appears, 
and the Hercynian limestone; this corresponds to the gap beneath the 
Baltic transgression. 

In the same way we see in England and Scotland various regions of 
the Old Red sandstone ; then more to the south, in the fragments of the 
folded Armorican and Yariscan ranges, the Devonian system, of great 
thickness and very diverse character ; consisting in Devonshire, the 
Ardennes, on the Rhine and in the Harz, the Thuringerwald, and the 
Sudetes, of quartzite, slates, and limestone in various alternations, it 
reappears, with many similar characters, linear Qraz on the border of 
the eastern Alps. In Bohemia the quartzites are absent and the thick- 
ness much diminished; the Hercynian limestone rests on the Silurian 
limestone. Within the eastern Alps also, in Styria, Carinthia„ and 
Camiola, the only representative of the Devonian system so far reoog* 
nized is the Hercynian limestone ^. 

In these regions, therefore, the Hercynian transgression does not, as 
in Russia, afford definite proof of its existence; stiU, here also, in spite 
of all subsequent disturbances, we plainly perceive from the distribution 
of the sediments that the littoral formations predominate in the north, 
and the pelagic in the souths 

We may now emphasize the fact that, according to the marine remains 
which have been found in the dolomite and limestone beds of Livonia and 

' A. Graf Keyserling and P. ▼. Elrueenstem, WisBenBchaftliche Beobachtongen anf 
einer Beise in das PetschorarLand, 4to, St. Petersburg, 1846, pp. 896 et seq. 

* G. Stache, Ueber die Silurbildungen der Ostalpen, mit Bemerkangen fiber die 
Devon-, Carbon- und Perm-Schichten dieses Gebietes ; Zeitschr. deutsch. geol. Ges., 1884, 
XXXVI, pp. 277-378. 

" 'The thick lower Devonian slates and grauwaoke deposits of the Rhine, France, 
and Spain, in spite of their imposing development, represent only a local forma- 
tion in western Europe, like the Trias formation of Germany, and indeed their petro- 
graphical characters as well as their poverty-stricken uniform faun{^ in which 
Gephalopods are almost completely absent, show that they must be regarded as thaUoto- 
sea deposits. It was always to be expected that it would be possible in time to 
point to the deep-sea equivalents of these shaUouhUHUer deposits ; as such the Hercynian 
series of the Harz and of Bohemia must be regarded.' A. Kayser, Die Fauna der lilte- 
sten Devonablagerungen des Harzes, p. 288. For the sake of completeness it may be 
observed that, in the Alburs, Tietze met with the red sandstone beneath marine Devonian 
deposits ; no organic remains were found ; Jahrb. geol. Beichs., 1877, XXYII, p. 389. 
Opinions as to the littoral or lacustrine formation of the red sandstone have been 
discussed by Gqd win- Austen, Rep. Brit. Assoc. Adv. Sci., 1869, p. 88 et seq. ; Ramsay, 
Quart. Joum. Geol. Soc, XII, p. 88, and XXYII, p. 241 ; Rupert Jones, A Monograph 
of the Fossil Estheriae, Palaeogr. Soc. ; Hull, Quart. Joum. Geol. Soc., XXXYI, pp. 252- 
274 ; Gosselet, Bull. Soc. g^ol. de Fr., 3« ser., I, pp. 409-417. 



CH.V] DEVONIAN TEANSGBESSION IN CANADA 231 

Conrland, the mazimnm of transgression in these countries must be placed 
in the middle Devonian (Stringocephalus horizon). At this period the 
Devonian sea attained its greatest extension in Russia. Immediately after, 
both in respect of age and distribution, follow the upper Devoiiian Domanik 
shales of the Petchora region (Goniatite shales) ; then marine formations 
are seen no more until the beginning of the Carboniferous period. 

Bearing this result in mind let us turn our attention to North America. 
Kayser has shown that the ' lower Helderbeig group ' of New York is the 
equivalent of the Hercynian stage. We need not stay to dwell on its 
limited distribution, the detailed classification of the succeeding series, 
or the complicated relations which arise in New York State from local 
unconformities \ We will confine our attention to one particular group of 
the Devonian beds of North America. 

The middle Devonian is represented in New York State by the Marcellus 
shale, the Hamilton group, and the Qenessee shale. The Marcellus shale 
below and the Qenessee shale above present so many points of resemblance 
as regards their fossils that Williams regards the Qenessee fauna as a 
recurrence of that of the Marcellus shale. The intervening Hamilton 
group includes beds of limestone with numerous marine fossils, and one 
in particular, the TuUy limestone, irregularly distributed towards the 
snnmiit, contains Rhyncoriella cuboides, a well-known middle Devonian 
Bpedes found in England, Russia, and on the Rhine '. 

In the Qenessee shales of New York State, close to lake Ontario, 
considerable quantities of bituminous matter occur, and Clarke has described 
a particular bed, formed of millions of shells of Styliola fiseiirella, a little 
Pteropod, scarcely 1^ to 2 millimeters long, which were washed up on the 
coast by the waves of the Devonian sea ; among them lie trunks of Lepido- 
d^idron, Dadoxylon, and other trees which were stranded at the same time. 
Sunilar beds of Styliola also occur in the Marcellus shales ^. 

These three members of the American series are universally recognized 
as forming the middle Devonian of Nqrth America ; they are the equiva- 
lents of the dolomite and limestone beds of Livonia and Courland. Above 
them lies the Naples shale, also containing bituminous beds in places, with 

' For these complicated relations, cf. in particular J. Hall, Proc. Amer. Assoc, 
1888, p. 66, and W. M. Davis, The folded Helderberg Limestones East of the Gatskills, 
BolL Mns. Comp. 2iOol. Cambridge, 1883, Geol., ser. I, pp. 311-829, pi. A local uncon- 
formity already disturbs the relations of the upper Silurian Niagara Group with the 
lower Silurian. 

' H. S. Williams, The Recurrence of Faunas in the Devonian Rocks of New York, 
Proc. Amer. Assoc, 1881, XXX, pp. 186-191 ; another recurrence of the same fauna 
is even asserted to occur in the Utica slates ; also by the same, On the Fossil Faunas 
of the Upper Devonian along the Meridian of 70** 80', Bull. U. S. Geol S^irv., 1884, 
No. 3, pp. 55-86. 

* J. M. Clarke, On the higher Devonian Faunas of Ontario County, New York ; 
op. cit., 1885, No. 16, pp. 41-120, pi. 



232 PALAEOZOIC SEAS [pabt in 

marine shells, fish remains, and terrestrial plants. Clarke found intercalated 
with these a bed of nodular limestone with Goniatites, which enables us to 
correlate this member with the base of the upper Devonian of Europe ; and 
at the same time it corresponds precisely with the petroleum^bearing 
Domanik shales of the Petchora ^. 

Thus the chronological equivalents of the Russian transgression on the 
other side of the Atlantic Ocean have now been determined. 

But it is precisely these equivalents of the maximum of the Russian 
transgression, L e. the rocks of the Hamilton group, which have been 
met with on the western border of the Canadian shield and in the valley 
of the Mackenzie, extending far beyond the region of the Silurian forma- 
tions of the United States up to the shores of the Arctic Ocean (II, p. 38). 
We have seen that amongst all the fossils collected between the Clear-wator 
in lat. 56* 80' N. and the Arctic Ocean, Meek recognized none but species 
of the Hamilton group. Above the calcareous Hamilton beds strictly 
so called, lie petroleum-bearing shales with beds of Stylicla fisvwreUa, 
a continuation of the Oenessee shale of the south, and the organic remains 
of these beds remain essentially the same through nearly thirty degrees of 
latitude, from Rock island, Illinois, up to the Arctic Ocean ^ 

Nowhere within the vast region which extends from the Clear-water to 
the Arctic Ocean is the Silurian known to occur ; future explorers may 
perhaps discover it. Tet it is hard to understand how deposits elsewhere 
of such wide extension, forming part of a flat-bedded series and lying near 
the Archaean foundation, should escape observation, and again, why the 
same members of the Devonian, and these only, should always have 
attracted the attention of the observer. 

In the present state of our knowledge we must conclude that the 
Hamilton beds, together with the overlying Genessee shales, advance in 
a great transgression over the western part of the Canadian shield ; thus 
they form the western glint and insert themselves on the north into the 
Silurian-Devonian-Carboniferous region which forms the Arctic archipelago. 
In what way these beds are continued within the zone of sandstone which 
lies in this region beneath the Carboniferous sandstone is at present 
unknown. 

These transgressive deposits of western Canada are, as we have said, the 
chronological equivalents of the transgressive deposits of the Russian 
Devonian. Thus we arrive at the condueion that a very cormderahU 
extension of the Devonian aeaa took place simultaneoudy from the Ural over 

^ The Naples shales occupy the lower part of the Portage Group ; F. Boemer refers the 
Goniatite beds of Bfldesheim and of Torbay in Devonshire to this horizon ; Lethaea, 
1880, I, pp. 50, 53 ; Clarke, torn, cit, pp. 38, 39, 49. Cardiola retrostriata^ Buch, extends 
through the whole of the Hamilton group and attains its maximum in the Goniatite beds. 

> F. B. Meek, Trans. Chicago Acad. Sci., 1868, 1, p. 77. 



CH.V] THE CAEBONIFEKOUS SYSTEM 

the Ruenan plain towards the weat and nortk-weat, and from tl 
mouiUains acrom the valley of the Mackenzie to the east. The oc 
dence eztands so far that the Domanik shales on the upper Petcl 
the contemporaneoTis Geneasee shales on the Athabasca are both chan 
by the preeraice of petrolenm. 

The poeitiTe phase in the middle of the Devonian s;^it«m thus g 
itself on both sides of the Atlantic Ocean at the same time. 

4b The CarboniferouB gyetem. In discnssing the phenomena m 
with the formation of the great coal deposits I shall again confin 
for the present to the North Atlantic region, using this term in tb 
sense, to indicate that part of the northern, temperate zone which 
from the Ural in the east to the Bocky momitains in the west 
nearly half a circle of latitude. In this region, thronghout tb 
of the Carboniferous, so striking a correspondence prevails betv 
meet important members of the series on both aides of the AtUu 
we may venture to consider Uie Enropean and American areas 
comparing them step by step. 

(a) The sediments of this period begin for the most part with a s 
depomt of variable thickness. In Scotland it is called the Calciferc 
stone, in Ireland the Coomhola grit, in eastern Canada the low 
measures, in Fenssylvania, according to the nomenclature' empl 
KogerB, the Vespertine and Umbral series ; towards the west th 
decrease in importance ; in Ohio they are known as the Waverley sa 
In Illinois, however, the clastic sediments derived from the land i 
and dolomitic limestone with marine fossila takes their place ; ti 
is then known as the Kinderhook group. 

Just as the series thins out as it proceeds from east Canada and 
towards the west, so it does in Europe when traced from Great E 
the east ; and in the south of Russia there reappear in the lowes 
beds of the Carboniferous some of the characteristic marine fossi 
Kinderhook group of Illinois. 

In Great Britain the remains of terrestrial plants, and in a f 
of marine animals, are interspersed in this lowest series. Kirkby 1: 
an exact account of the succession in the Calctferous limestone 
in Scotland. Here the lowest part of the series is not exposed ; i 
less we are able to examine a thickness of more than 8,900 feet; 
the npper 8,400 feet — that is, the whole series with the exceptit 
lowest part — there are about fifty thin beds of coal three inches and 
in thickness, many of them resting on beds of clay wiUi Stigmari 
are regarded by most observers as the original roots of the forest 
or as a proof that the coal was formed m aitit. Plant remains a 
scattered through the sandstone. In addition, there are not 1 
eighteen beds of shale or thin layers of limestone intercalated in tl 



284 PALAEOZOIC SEAS [paet in 

at various levels and containing marine shells. Thus, 2,280 feet below the 
upper limit of the Calcif erous sandstone an Encrinite bed occurs with 85-^ 
species of marine shells, and numerous Crinoids, and this contains stems 
of Lepidodendron and Dadoxylon \ But if the underclay with Stigmaria 
is to be recognized as the original' soil, then the conditions favourable to the 
growth of land or marsh plants must here have alternated often indeed 
with periods of occupation by the sea. 

On the Atlantic coast of Ireland the thickness of this series amounts, 
according to Hull, to about 1,500 feet ^ 

In New Brunswick and Nova Scotia, where the series attains a 
prodigious thickness, the occurrence of workable coal is well known, 
but intercalations of marine beds are absent In the west, on the other 
hand, in Illinois, the Einderhook group is formed chiefly of friable beds, 
often dolomitic, which, as we have seen, contain marine fossils ^. 

In the Pyrenees and in Asturias the 'Marbre griott^' distinguished 
by its Ooniatites must be assigned, according to Barrois, to the lowest parts 
of the Carboniferous system ^. 

(6) Above this lowest division of the Carboniferous there lies, in Europe 
as in America, the Carboniferoria limestone, which, by its abundance in 
all kinds of marine animals and by the great thickness it attains, indicates 
the prolonged existence of the open sea over the greater part of the region 
before us. We can scarcely otherwise represent the facts than by supposiDg 
that after the singular oscillations which took place in Scotland during the 
deposition of the Calciferous sandstone, just illustrated by an example, 
pelagic conditions were established far and wide. 

In England, according to Hull, the Carboniferous limestone increases in 
importance towards the south-east ; in Belgium it occurs with a thickness 
of 800 meters ; it extends over a lai^e part of Germany without acquiring 
any great thickness, and then broadens out in a great sheet over the Russian 
plain, extending far beyond the Ural and into the Arctic regions. It forms 
a large part of the surface of Ireland, attaining there a thickness of 2,500 
to 8,000 feet, and it occurs in France as well as in Spain. 

Across the Ocean it appears in eastern Canada, with intercalations 
of reddish shales accompanied by gypsum, an indication of interrupted 

^ J. W. Eirkby, On the Zones of Marine Fosflils in the Galciferona Sandstone 
Series of Fife, Quart. Joum. Geol. Soc., 1880, XXXYI, pp. 559-590 ; also B. Etheridge, 
jun., On our present Knowledge of the Invertebrate Fauna of the Lower Carboniferous 
or Calciferous Sandstone series of the Edinburgh neighbourhood, op. cit., 1878, XXXIV, 
pp. 1-26, pi. 

' £. Hull, Physical Geography and Geology of Ireland, 8vo, London, 1878, p. 31. 

* A. H. Worthen, Geology of Illinois, 1868, III, p. 115 et passim. 

* C. Barrois, Becherches sur les terrains anciens des Asturies et de la Galicie, 
Mem. Soc* G^ol. Nord, Lille, 1882, I, pp. 570, 576, and 588 et passim ; also Bol. Com. 
Mapa geoL EspiAa, 1881, VIII, pp. 131-155. 



r" 



CB. y] CULM AST) MILLSTONE GKIT 

depoeition and the temporary existence of evaporfttdng U^ 
absent in Yirginift, where a great deposit of sandstone tak 
Beyond tiiis region to the west, on the other hand, it begins 
Bgun in thickness. In the soath-west of lUinois it is 1,200 i 
thick. From Qua we see that the Carboniferotu limestone ol 
part of the United States was separated from the Atlanti 
a deposit of sand, that is by a clastic zone occupying very 
site of the existing mountain chains of the east, and that ( 
oommunication with the Carbonif erooa sea of Europe could onl; 
place by a circnitoos route. Nevertheless there is a close coi 
between the faunas of the two regions, and with the progress i 
tkm the Dumber of identical species continually augments '. 

Several series have been distinguished in the Carboniferon 
by Hall and Worthen in America, and by Qosselet and Dnpo: 
and Belgimn. Do Koninck has made an attempt to compart 
one another, as well as with those of Russia '. 

(c) and (d) In Europe a stage now appears which has not b 
in America; in typical locidiUes it presents the character < 
formation with a mixture of terrestrial plants and marine i 
IB the CtdTn or the Toredale beds. Its flora is widely distribi 
tlie limits of Eorope. The Culm occurs in Ireland and England, 
on the Rhine, in the Vo^es, and in the Harz, and following i 
folds it extends, not indeed without interruption, but with cons 
ters, as far aa Moravia. It is known in France, F. Roemer hae 
it occurs in Spain and Portugal, and Toula has discovered it in 



Some observerB hold the opinion tiiat close relations es 
the Culm and the v/nproducHve aandatone or MiilstoTie grit ; 
deposit which occurs in some regions in great thickness ai 
in others. Where the Culm is not present the unproductiv 
tests directly on the Carbomferous limestone. From Htill's usel 
we see that the Millstone grit and beneath it the Yoredale bee 

* Dawion, Acadian Geolog7, in particular p. 276 et aeq. 

* e. g. 8. 0. Perceval, PalaeaeU atneata, 6eoL Mag., 1876, 2nd ser., Ill, p. 2 
and Nicholson, On Palaeacis, Ann. Mag. Nat. HiiL, 1878, 6th ser., I, i 
Q. A. Lebonr, Note snr denx foasiles da Calcaire carbonif^ de Northnni 
8oc. gtel. Belg., Liege, 187S, III, p. 21 et rnq. 

' L. 0. de Eoninck, Note nir le Spirifir MotqutntU ; Ball. Mdb. Hilt, ne 
n, pp. 371-379. 

' F. Boemer, Uebei da* Yorkommen TOn Cutm-Schicliten mit i^MuIonooi 
dem Sadabhange der Sierra Morena, Huelva, Zeitechr. deutsch. geoL Oei 
pp. 589-592; and b; the same. Debar das Torkonunen von Culme-! 
PiuidoHomga Bedten in Portngal, torn, cit., 1876, XXTIII, pp. 8&4>3 
Geologucbe UnterBochnngen im weatlichen Theile dea Balkan, Siti. I 
Wien, 1878, LXZVII, pp. 249-317, and in particular pp. 253 and 307. 



286 PALAEOZOIC SEAS [pabt ni 

greatest thickness in south Lancashire ; indeed the first is estimated 
at 8,500 to 5,000' feet, the second at 2,000 to 4,000 feet. Li Torkshire, 
Derbyshire, and north Staffordshire the thickness is still considerable; 
then it diminishes in all directions somewhat rapidly, and only near Bristol 
rises again exceptionally to about 950 feet \ 

The unproductive sandstone reappears in Westphalia and there rests on 
the Culm ; von Dechen gives confirmative evidence to show that these two 
deposits are difficult to distinguish, and several species of Culm plants 
appear in the unproductive sandstone K 

Beyond the Ocean we see the unproductive sandstone, in Pennsylvania 
1,000-1,500 feet thick, extending along the Appalachians towards Virginia 
and Tennessee, and at the same time increasing rapidly in thickness towards 
the west. The occurrence of the Carboniferous limestone in Europe as 
in America, succeeded in both regions by the unproductive sandstone, is 
so striking a fact that it was adduced by Dana as a conspicuous proof of 
the contemporaneity of the changes which have affected both hemispheres '. 

(e) We have now reached that series of deposits in which the greatest 
quantity of fossil fuel lies accumulated. Many unsolved problems are con- 
nected with the origin of the coal, and it will be necessary to enter into 
some detail. 

Marine conditions are on the wane, but have by no means altogether 
disappeared. In the mighty deposits composed chiefiy of sandstone and 
shale, that is> clastic sediments derived from the land, the coal beds as a rule 
are repeated one above the other a great number of times. The coal-field 
of Ostrau and Karwin in Moravia and Silesia includes two series of ooal- 
bearing formations of different age. Leaving out of account the beds 
under 15 cm. in thickness, the lower divisions contain 179 coal beds in 
a thickness of 8,798 meters ; and the upper division, which is 415 meters 
thick, includes thirty-nine coal beds; or both together 218 coal beds in 
a formation 4,208 meters thick, and on an average one meter of coal to 
twenty-eight meters of sandstone and shale ^. 

In contrast to this we see that in central Bohemia the coal-bearing 
formations contain coal only in their lowest part, and the number of beds 

^ E. Hall, On the Upper Limit of the essentially Marine Beds of the Garboniferoofl 
Group of the British Isles and adjoining Continental Districts ; Qnart. Joom. GeoL Soe^ 
1877, XXXIII, pp. 618-650. 

* Von Dechen, Erl&nterongen znr geologischen Earte der Rheinproyins und der 
Provinz Westfalen, II, 1884, p. 220 ; D. Stur, Verh. k. k. geoL Reichg., 1876, p. 266. 
Also Hull (torn, dt., p. 619) points out similar relations between the Culm and the 
sterile sandstone. 

' J. D. Dana, Manual of Geology, 2nd ed., 1875, p. 894; also H. Martin Chance, The 
Millstone Grit in England and Pennsylvania, Am. Journ. Sci., 1881, XXI, p. 184. 

* Monographie des Ostrau-Earwiner Steinkohlen-Beyieres, herausgegehen und bear- 
beitet vom Berg.-Htitt. Yer., M&hr.-Ostrau, 4to, Teschen, 1885, Geognostischer Thefl 
▼on W. JiSinsky, p. 18. 



CH. y] UNDERCLAY 

over 15 cm. Hhick hardly amounts to mors than eight or ten, of v 
the lowest it is true attain an impoeing thickness. These meaeuref 
folloved hy a very great accumulation of unproductive sandstone 
shales, and then by upper Coal-measures, perhaps of Permian age '. 

The Coal-measures of Moravia and Silesia rest conformably on 
Colxn, and a part of the Culm flora ext^ids into their lowest beds ; 
Coal-meaaures of central Bohemia, on the other hand, rest in transgre 
on Archaean and Silurian rocks. The first region contains marine i 
calations, the second does not. This is the first example which shows 
manifold were the conditions under which the Coal-measures were prodi 

Marine beds are not seldom intercalated with the Coal-meae 
It has often been nudntained that the repeated formation of coal bed 
been caused by so many oaciUationfi of the solid land. In England espec 
insiBtence has been laid on the fact that below most of the coal beds a 
the ' underday,' is present, filled with the still rooted trunks of the p 
which formed tiie coal, as we have already mentioned in speaking o: 
Calciferous sandstone of Fife : the eoal most thus have been forme 
place, and each particular bed indicates the occurrence of a fresh 
surface or marsh, on which a new forest growth arose. We wiU 
now inquire how often the whole solid crust of the planet most have 
moved upwards and downwards in order to form such series of 
but will first attempt, b^inning in the west, to obtain a general so 
of the varioua types of development among the Coal-meaaures. It will 
be seen that above the marine fauna of the Carboniferoos limei 
mentioned above there exists in the Culm and in the intervening 
another, although not sharply differentiated, marine fauna, and that 
is followed later by yet another fauna, tiie marine fauna of the d 
Carboniferous or that of the Fuauliua limestone. 

In Utah and Nevada, the most westerly part of the regions consit 
here, the whole Carboniferous is represented by muine deposits -^ 
attain an extraordinary thickness j Clarence King gives for the Wahi 
limestone, which contains lower Carboniferoos fossUs and perhaps exi 
down into the Devonian, a thickness of 7,000 feet in the Wahsatcb m 
tains ; for the Weber quartzite resting on the limestone 6,000 feet, ani 
the upper CarboniferoTis limestone 2,000 feet. The figures given for ce 
Nevada are not quite so large. It is worthy of note that in Nevada, ii 
midst of masses of marine limestone, amongst corals and brachio 
Waleott has discovered two species of air-breathing gastropoda which 
evidently have been borne in from a distance '. 

* 11. V. lipoid, Du Steinkohlengebiet im nordwettlichen Theile dei Prtiger E 
in BOhmen ; Jahib. k. k. geoL Reichi., 1661 &nd 1862, XIT, pp. 481-525, pi. 

■ JF^jwa pri»ea, Zi^*jfekiiu cattoHoria. Wftloott, Palaeontology of the Eureka Di 
1884. Tin, pp. 262, 263. 



238 



PALAEOZOIC SEAS 



[PABT m 



To the east, on the outer border of the Bocky mountains and as far as 
the Black hills, Dakota, the thickness greatly decreases ; but the deposits 
are all marine. 

Still further to the east beyond the prairies, the Coal-measures b^in to 
make their appearance. At first, in eastern Nebraska^ the coal beds accord- 
ing to Meek are certainly very rare, small and scarcely workable ; beds of 
marine Fusulinik limestone alternate with beds of shale which include the 
scattered remains of half -decayed plants and some tree stems ^. 

Beyond the Missouri in Iowa there are many workable seams, but always 
associated with the marine limestone. The Coal-measures rest on the lower 
Carboniferous limestone. In the south-east, where there was open sea^ the 
Carboniferous limestone is everywhere visible beneath the Coal-measures, 
but towards the north, in Iowa and Illinois, the latter extend far beyond 
the Carboniferous limestone, and often rest on Devonian or Silurian beds. 
Towards the Appalachians the Carboniferous limestone is represented, as we 
have seen, by sandstone, conglomerate and shales. During this transgression 




Fzo. 25. Sedion n^ar Iowa ciiy, after J. HalL 

a, Deronian limestone ; h, coarse sandstone in wairy layers ; c, gre j and greenish elay ; d, 
small seams of coal, the lower part atiXj^ containing fish teetti. 

it happened that the transgressive coal-bearing sediments penetrated from 
above into cavities of the Devonian limestone, and formed within these 
cavities little stratified deposits which even contain fish remains, and in one 
place a little bed of coal. These facts, which have been described by J. Hall, 
have an important bearing on explanations to follow ' (Fig. 25). 

The Coal-measures of this region are distinguished by the frequent 
intercalations mentioned above of f ossiliferous marine beds between the coal 
seams ; Worthen describes a section at New Haven on the lower Wabash, 
Illinois, in which twenty f ossiliferous marine intercalations and sixteen ooal 



^ F. B. Meek, Report on the Palaeontology of East Nebraska (in Hayden, Rep. XT. 8. 
Geol. Surv. Nebraska), 4to, Washington, 1872, p. 1S4 et passim. 

' J. Hall, in HaU and Whitney, Report on the Qeological Surrey of the State of Iowa, 
8to, 1858, pp. 117, 130, 181. Richthofen saw the same thing in Shantung and Sbami; 
of. his China, II, pp. 203, 411, 437, 718; and in Newfoundland Garboniferona fossils 
are found in the cavities of the Silurian limestone ; Murray and Howl^, Qeological 
Report, Newfoundland, London, 1881, p. 333, note. 



CH. v] COAL-MEASURES OP BRITISH ISLES 239 

beds oocor in alternation. Some of the marine beds attain a thickness of 
thirty-five feet \ 

These marine intercalations continue to Ohio and west Virginia; in 
Pennsylvania they come to an end. They are also absent in eastern Canada. 
In the section of the south Joggins, on the northern arm of Fimdy bay, 
which was studied by Logan, Lyell, and Dawson, the Carboniferous forma- 
tion is exposed alon^ the coast ; there are eighty-one beds of coal, most of 
them resting upon underday. We meet with stems of trees standing 
upright, reptiles in hollow tree-trunks, air-breathing moUusca> remains of 
fishes also, but no marine intercalations \ 

In continuing this rapid survey of Europe we must now distinguish two 
groups of Coal-measures. The first belongs to the outer border of the 
Annorican and Yariscan chains, or lies to the north of this border. It 
comprises all the coal-fields of the British isles and of the north of France, 
next those of Belgium and Westphalia, and then, after a long interval, those 
of upper Silesia and Moravia. The second group lies south of the first; the 
coal-fields on the Saar and in central Bohemia, and that of Schatzlar- 
Waldenburg, belong to it. In the first group marine intercalations are 
known to occur, in the second there are none or at most feeble indicationa 

Where the first group forms a part of the border of the folded ranges it 
rests conformably on the next older deposits. The second group lies trans- 
gressively on various rocks. This at least appears to be the rule from 
Moravia onwards as far as the Rhine. 

The numerous isolated basins which contain the rich Coal-measures of 
the British isles are probably only the remains of a once continuous forma- 
tion, broken up by denudation and preserved owing to their tectonic 
position. In the important coal-fields of the south-west of England, i. e. 
those of South Wales, the Forest of Dean, and Bristol, the original con- 
tiQuity may be most clearly recognized. The coal-fields of Coalbrookdale, 
south Stafibrdshire, and Warwickshire are still perhaps connected imder- 
ground. The coal-fields of Ireland and of the great trough of Scotland 
may also be regarded as parts of a single sheet. At the same time a very 
great degree of variability characterizes the thickness of the imderlying 
coal-bearing sandstone, as well as the nature of the overlying beda 
The coal-field of south Staffordshire rests unconformably on the upper 
Silurian *. 

Marine intercalations are found only in the lower parts of the Coal- 
measures; higher up they become rare, of trifling thickness, and present 

^ A. H. Worthen, Geological Survey of Illinois, 8vo, Boston, 1875, VI, p. 2-5. 

* Dawson, Acadian (Geology, pp. 15(^218. The whole thickness is quoted on Logan's 
authority at 14,570 feet, but it is doubtful whether the uppermost of these beds do 
not belong to the Permian period ; Dawson, Quart. Joum. Geol. Soc 1874, XXX, p. 209. 

' A. C. Ramsay, The Physical Geology and Greogiaphy of Great Britain, 5th ed., 8to, 
London, 1878, pp. 119-128. 



240 PALAEOZOIC SEAS [pabtih 

a purely littoral character ; still higher they disappear altogether. In a 
general account of these intercalations by Hull, the lower subdivisions are 
spoken of as the Oannister beds \ 

The Qannister beds, characterized by the presence of marine fossils, are 
present in all the Irish coal-fields, with the exception of the BaUycastle 
coal-field in Antrim, which belongs to a still lower horizon : they also occur 
in Scotland, and in the various basins in England, whence they are con« 
tinned into the continent. The same marine intercalations are found in the 
Coal-measures of the north of France, as for instance near Auchy-au-Bois, 
and at several localities in Belgium. In the basin of Charleroi the gaping 
valves of a Mytilua still attached in pairs occur in great numbera Thisis 
precisely what may be seen on a modem beach when mussels have quietly 
decomposed : the adductor muscles give way first, then the shell flies open. 
These beds thus mark undisturbed deposition \ 

In the Coal-measures of Westphalia there are numerous intercalations 
containing shells; the lower of these are of purely marine origin, but higher 
up they contain only species of the genus Anthracosia, the marine nature of 
which is, to say the least, doubtful. Von Dechen has given the complete 
succession of these beds \ 

The presence of marine intercalations in the Coal-measures of upper 
Silesia was first discovered by F.Boemer in 1868, and in 1870 he compared 
this remarkable series with the Qannister beds or ' Pennystone ' of Coal- 
brookdale and Carluke in Scotland. Eosmann has shown that they 
repeat themselves at definite horizons and may render great service in 
determining the succession of the coal bed& The shell-bearing beds are 
partly of marine origin, doubtless indeed littoral, and contain such genera 
as Phillipsia, Bellerophon and Productus ; others exhibit limnic characters, 
and these contain Anthracosia and Modiola \ 



^ Hnll, Upper limit, ftc, p. 616 et seq. 

* C. Barroifl, Notice sur la fauna marine dn terrain honiller da baasin septentrional 
de la France, BnlL Soc. g^ol. de Fr., 8« s&r., II, 1878-1874, pp. 223-226 ; GosMlet, 
Esquisse g^ologique du Nord de la France, 8vo, Lille, 1880, p. 149 et seq. ; Briart et 
Comet, Notice sur la position stratigraphique des lits coqnilliers dans le terrain houiller 
du Hainaut, Bull. Acad. Roy. Belg., 1872, 2* s^r., XXXIII, pp. 21-81 ; R. Malherbe, 
Des horizons coquilliers du syst^me houiUer de Li^ge, Ann. Soc. G4ol. Belg., Liege, 
1876, III, p. Ixvii et seq. Beds are here considered which, judging from the character 
of the fossil shells, are not indubitably of marine origin; the gaping Mytilus are 
described by G. Blanchard and J. Smeysters, Note sur quelques fossiles rencontr^ 
dans le syst^me houiller de Charleroi, op. cit., 1879-1880, VII, M^m., p. 15. 

* H. von Dechen, Erlftuterungen zur geologischen Earte der Rheinprovini, II, 
p. 247 et seq. 

* F. Roemer, IJeber eine marine Conchylienfauna im productiven Steinkohlengebirge 
Oberschlesiens, Zeitschr. deutsch. geol. Qes., 1868, XV, pp. 567-606; by the same. 
Geologic Ton Oberschlesien, 8to, 1870, in particular pp. 94, 95 ; Weiss, Zeitschr. deutsch. 
geol. Qes., 1879, XXXI, p. 219, Sec; Eosmann, Die neueren geognostischen und 



r] COAL BEDS OF OSTRAU 241 

he investigatioiis puiSned for many years by D. Stur, on thb prolonga- ) 

of Ute apper SUesian coal-field into Austria at Ustran futd Earwia, 
established the following facta. The Coal-measures lie conformably 
e.Cnlm, which contains terrestrial plants and at the same time marine 
ds, as in Naasau. A part of the fiora of the Culm ascends into the 
' division of the ooal-bearing beds, artd forms together with additional 
s the upper Culm Bora, the zone of Sj^enopkyllv/m, tenerrimum, or 5 

strau beds. This zone includes five groups of coal seams ; in the three 
tower of these marine intercalations occnr similar to those observed by 
F. Roemer in npper Silesia. The most important of these lies between the 
third and fonrth coal-bearing groups ; Stnr likewise believes that it repre- 
sents the English Qannister. In the fourth and fifth groups we have only 
the genus Modiola, then Anthrscosia in great quantity, and finally Eury- 
ptems, Cypris, and Planorbis. The marine character has disappeared. 

Now follows a higher division of the ooal-bearing formation with a new 
flora, the Schatzlar beda ; in these, marine intercalations are unknown ^ ] 

These facts show that throughout the central part of the United States, I 

from Indiana and Iowa to western Pennsylvania, frequent marine inter- i 

ealations occur between the coal beds, and these only cease in those eastern 
rc^ons where, as in the Alleghanies, the underlying Carboniferous Umestcme 
is also replaced by sandstone and shales ; or where, as in New Brunswick, 
the proximity of the shore is revealed by int«rbedded gypsiferous marls. 
The intercalations are more considerable in the lower parts of the coal- 
bearing formations, and there the marine origin is more clearly manifest, ; 
while towards the upper part littoral characters become increasingly evident ' 
and the intercalations are in general of slight importance. 

Similar marine intercalations are present in the Coal-measures of I 

Scotland, Ireland, England, the north of France, Belgium, Westphalia, I 

upper Silesia, and north Moravia. Here also the lower beds contain j 

a fauna certainly marine ; these are followed, either immediately or after j 

B(Hne previous alternations, by littoral mollusca or by mollusca of uncertain, 
perhaps of lacustrine origin, such as the Anthracosia ; still higher in the I 

aeries, however, these intercalations are wholly absent. j 

Notwithstanding this apparent correspondence the intercalations in | 

Europe and America are not of the same age. Those of Europe are the I 

Qannister beds, which are closely allied by their fauna to the lower 
Carboniferous ; certain new species contribute an additional feature and 
the more important groups characteristic of the open sea are usually absent. 

pilfioiitologiieheti Anfitctlllaae nnf der Konigsgrube bei EOnigshQtte, Ober-Schleaien ; 
Zeitachr. f. B«rg.-, HQtt.-, Salin., Berlin, 1880, XXVIII, pp. 305-340, maps. 

* D. Star, Die Calmflora der Ostranec und Waldenburger Schicbten ; Abb. k. k. 
geol. Raichs., 1877, VIII, p. 428 et «eq. According to Stur the maiine intercalations 
ben belong to Coal-meaniea lower in the uriei tban tbose of Weatphalia &nd Bel(pmii, 
wbicb he uoigni to the Schatzlar group. 



242 PALAEOZOIC SEAS [PABTin 

The American intercalations, on the other hand, contain the fauna of the 
upper Carboniferous, that is of the Fusulina limestone. 

The case is otherwise in southern Europe. 

In Asturia8,according to Barrois,the Carboniferous limestone is succeeded 
by beds containing the Culm flora, these by an alternating series of beds, 
some containing plants and others marine moUusca. which ascend up to the 
level of the intercalations met with in America on the horizon of the 
Fusulina limestone, as is shown by the identity of a large niunber of species K 

Still more remarkable are the facts observed in the southern Alps. On 
the Auemig and the Eronalp, near Pontaf el in Carinthia, we find a frequent 
alternation of yellow sandstone, which contains plants belonging to the 
uppermost Carboniferous flora, and of dark limestone filled with Fusulina 
and characterized by a marine fauna, which may be correlated with that 
of the upper Carboniferous intercalations of America. The sandstbne is 
associated with beds of quartzose conglomerate ; sometimes a few isolated 
marine shells are found along with its plant remains \ 

There is thus a complete correspondence between these sections in 
Carinthia and those of the coal-fields of Illinois and Iowa. In Carinthia 
the coal beds are represented, by plant-bearing sandstone, but the alternating 
f acies is present in one locality as in the other. 

In the south of Russia workable measures crop out on the Doneta. 
Some difference of opinion prevails as to the succession of the strata in the 
principal region, but on the north border of the basin near Kaluga, Tula, 
and south of Rjasan, the facts have been completely elucidated by the 
investigations of A. Struve K 

' Barrois, Asturies, pp. 582, 593, &c. 

* I have frequently stayed for a connderable time in the Eflhweger and Watschiger 
Hatten and the Hdtte am Ofen in order to gain a knowledge of these beds. In 1879 
I traced the BQccession in the Krone ; it appeared that the little patch of Goal-measoiei 
with Prodnctus, which is also mentioned by Stache, is separated from the series by 
a dislocation. It lies higher than the yellow sandstone beds, which are distinguiahed 
by large specimens of Spirophyton, and shows an alternation of sandstone and qnartaoM 
conglomerates repeated four or five times, and of bluish-black Fusulina limestone. Plant 
remains occur in the sandstone, which D. Stur has been kind enough to determine ; they 
all belong to the highest division of the Carboniferous; among them are Annularia 
sphenophylloides and PeeapUrie longifolia. An occasional braohiopod may be found 
with them in the sandstone. In the Fusulina limestone, on the other hand, we meet 
with Phillipsia^ Conocardium, &o. Most striking is a highly ornamented gastropod, 
probably Naticopaia nodosa, Meek and Worfchen, or the variety N, Wbrihem, Barrois. 
Stache many years ago recognized the American character of the beds in certain 
localities of the southern Alps ; that he should take them for Permian was intelligible 
at a time when the limit of the Permian had been placed too low even in the Americas 
localities. 

' A. Struve, Ueber die Schichtenfolge in den Carbonablagerungen des ifidlicben 
Theiles des Moskauer Eohlenbeckens ; M^m. . Acad. Imp. ScL Saint-P^tersb., 1888, 
XXXIY, No. 6, 106 pp., map. The highly important distinction between the beds witk 
Spiri/era mod^uensis (Fusulina limestone) and those with Fh)dueH»s giganUuB (npper 



CH. v] COAL BEDS OF RUSSIA 243 

It may be well to recall, however, before considering this district, that 
very different effects may be produced by the same cause, according as 
it acts in the neighbourhood of the shore or at a point some distance &om 
ii. In the proximity of the land the clastic sediments may attain a thick- 
ness BO great as to bring them nearly to the surface of the sea, but further 
away an .accumulation on a scale equally grand is not to be expected, 
bat only a thin calcareous deposit. Further, a positive displacement, so 
trifling that the growth of the clastic sediments is able to keep pace with it, 
imiy give rise to a tract of marshy land always advancing both inland and 
aeawards, while at the same time in the open sea the same displacement 
will merely bring about an increase in depth. 

South of Moscow we again encounter the Fusulina limestone, but without 
intercalated coal seams or plant-bearing beds. Beneath it lies marine lime- 
stone with Productue gigardeus, the highest horizon of the Carboniferous 
limestone. Towards the base this marine stage includes two beds of lime- 
stone containing Stigmaria and occasionally passing into sand and clay with 
thin beds of coaL Beneath these lie the workable measures, and beneath 
the measures, marine limestone beds, in their lowest part containing, as we 
have already seen, a few fossils of the Einderhook group of Illinois. 

Thus in Bussia also the alternation of marine beds and Coal-measures is 
not wanting. The alternation, however, is less frequently repeated ; it does 
not occur except in the lower horizons, far below the Fusulina limestone. 

From these observations it appears that this kind of alternation is very 
eharacteristic of the Coal-measures of the Carboniferous system. We see 
it in the Calciferous sandstone of Fife, beneath the Carboniferous lime- 
stone ; then in the Qannister beds overlying the Carboniferous limestone 
and extending from England as far as Moravia; finally, in the upper 
Oaf boniferous horizon of the Fusulina limestone in the United States, 
probably in Spain, and certainly in Carinthia^ where the coal beds are 
represented by plant-bearing sandstone. 

Far away from the regions considered here, in the coal-fields of north 
China^ Bichthof en has met with the same intercalations of marine beds ; 
and there also the terrigenous clay of the coal-bearing series unconf ormably 
overlaps the Carboniferous limestone, and, as in Iowa, fills the cavities 
which have been excavated in the limestone by subaerial denudation; 
in south China, however, marine beds of the upper Carboniferous rest 
directly on the Coal-measures \ 

GarboniferoQS limestonQ) is also to be found in Eoninck, Bull. Mns. Belg., 1883, II, 
p. 871 et teq. In the lowest beds of the Carboniferous limestone of Russia, StruTe 
identifies several species as occurring in the American Einderhook group. On the 
Bnssian plant remains see Stur, Yerh. k. k. geol. Beichs., 1878, pp. 219-224. 

^ F. von Bichthofen, China, II, pp. 205, 717, 782 et passim ; p. 208, figure which 
shows the introsion of Carboniferous clay into the cavities of the underlying limestone. 
Cf. note 2 on p. 288. 

B 2 



244 PALAEOZOIC SEAS [pabt n 

Although many thousands of men work day and night in our Coal 
measures, and although many acute obaerverB are led by their profession ti 
make the study of these deposits the business of their life, yet the modi 
of formation of the coal beds is still far from being satisfactorily explained 

la England, geologists, as we have observed, lay great stress on th- 
onderclay with its Stigmaria, which they regard as the soil of a marahi 
forest, possibly growing out towards the sea, like the mangrove swamps o 
the present day ; they take the Stigmaria for the roots, and point to th< 
upright trunks of trees, three to four meters or more in height, which hen 
and there may be seen in the sandstone standing above the coal. ^Rii 
coal itself must consequently have been formed in the place where it ii 
found, and lies at the foot of the trunks and over the ttuigle of roots, likt 
the litter of an existing forest. 



Fro. 2S. HmH Fomtt r^ tht Carlanifirmu Period. At Trfive, near SalDt-Etienne, after 

Qrand'Eury. Vertical height about 13 tneten. 

/. Calatnila; a. Ptcronivt; j. SigOlaria; 4. Cordaitn; j. CalamodindnH, 

But the upright truoks prove that the sandstone cannot have beei 

slowly deposited. Later on we shall point out that on the shores of tin 

North sea peat bogs are to be seen lying some feet below the level of mean 

tide with the stumps of trees rising out of them ; further, that sand dnnef 

advancing inland, press down the peat below the level of the sea, and thai 

in the sand of the dunes, so long as it covers the peat, the trunks of th« 

buried forest may be seen, standing upright or more or less inclined 

Forests thus overwhelmed by the action of the wind were observed in th« 

Bermudas by the members of the Challengar expedition ; the trees an 

buried in fine calcareous sand, which hardens to a friable rock and encloset 

the remains of the trunks still standing erect. A slow oscillation of the 

strand-line would never bring about this result, the trunks would deca^ 

and disappear'. In some cases the phenomenon has been explained b; 

^ ThoDiRon and Hnmij, Namtive, I, p. 141, fig. 55 ; p. 142, fig. 56. It \a» ben 

■apposed th&t the trnnks had placed themselveB TeTticall^ while floftting and had 



FORMATION OP COAL BEDS 

g that as the sedimeot grew upwarda, the sterna of the plant 
roots laterally, bat even in this case the whole thickness 
; in question mnst have been deposited within the lifetime of 
The sketches of the sections to be seen in the open quan 
ieune and published by Qrand'Eury show clearly how the vegE 
rhe depoeition, and how fresh individuals again and again i 
iw beds. It is true Qrand'Eury expressly points out that 
iible growth of trees or roots is cut off above by a ' dess( 
ihough shaved off by a plane, above which the next bed 1 
h the succession of the forests ia so clearly displayed in the bi 
e, yet Qrand'Eury has been led by the consideration of the 
ces, and particularly by a thorough study of the process of ve| 
eition, to return to the view that the coal beds have not gn 
it are the remains of decaying planta transported by wat< 
1 in stratified beds one above the other K 
OS now glance once more at Fig. 26, p. 288, which represei 
1 of Carbonifeiooe sediments into a hollow of the Devoniai 
rhere is first a deposit of clay with fish teeth, and above this 
coal ; but we can only account for the formation of the coal I 
hat the whole cavity was once filled with muddy water cont 
eing vegetable matter in suspension which afterwards co 
tp of the cavity. 

definitely asserted that in Illinois the coal beds do not res 
ly, but directly on shale or limestone ; in these cases the trs 
»ble matter from some other locality must be admitted, a 
lat in Nebraska these coal beds pass into beds of shale, tl 
oly isolated fragments of vegetable d^ris are strewn ', 
e appears to be no doubt that thick beds of coal may som 
into a nxunber of smaller seams, which become separated i 
rher from one another by the thickening out of intercalated ^ 
a rock. 

England several examples of this are known. Jukes has 
outh Staffordshire the main coal bed, twenty 'five feet in thi< 
wards the north into nine aeama, in such a manner that the 
18 and the intervening beds amonnta to 890 feet. One thii 



I vertically into the ledimenta, bnt thia can hardly be true of whole 
, Sot t'origine dea tronca d'orbrea foaailes petpeadicnloirei mix iti 
)niUer, Compt Rend., 1881, XCIII, pp. 160-163. 

anu)d'Eni7, Flore cubonifere da ddpartement de la Loin et do 
ranee, 4to, PoriB, 1877, with Btlaa, pi. xxxiv, and MSmoire mr la 
» houille, Aon. Mines, 1882, 8< s^r., Uemoirei, I, pp. 99-292, pi. ; Dei 
.6. 

Worthen, Geolt^cal Surrey of IllinoiB, 1886, I, p. 70. 



246 PALAEOZOIC SEAS [pabthi 

parting swells out within a distance of one mile into a wedge of unpro- 
ductive measures 128 feet thick \ 

The great bed of Commentry (AUier) divides, according to Fayol, into 
six smaller seams which diverge farther and farther apart '• 

In the United States this question has given rise to very searching 
discussions. Andrews having maintained the uniform and continnons 
extension of the coal beds, it was shown by Newbeny and Stevenson that, 
as a matter of fact, certain thick beds, owing to the swelling out of the 
intervening masses, are divided towards the middle of the original basin 
into several seams sometimes widely separated from one another. At the 
period of the upper Coal-measures, i. e. during the formation of the chief 
beds of coal, the eastern region of North America had already been divided 
by the anticlinal of Cincinnati (I, p. 557) into two basins which nnite 
towards the south-west, sp that Ohio, Pennsylvania^ and west Virginia 
show a certain independence in respect to Iowa, Illinois, and the western 
regions, notwithstanding the correspondence in character otherwise so 
marked between the marine intercalations. It is towards the bottom 
of these two basins that the splitting up of certain of the marine seams 
appears to become complete ^ 

* I am therefore compelled to believe,' writes Stevenson, * that all the 
coals of the upper coal group are offshodts from one continuous marsh 
which existed from the beginning of the aera to its close, and which in its 
full extent is now known as the Pittsburg coal seam. During the whole 
time of the formation of the upper coal group the general condition was 
that of regular subsidence interrupted by longer or shorter intervals of 
repose. During the time of subsidence the marsh advanced up the side 
of the trough, as new land was continually becoming fitted for its support. 
During repose, deltas were formed in the bay, and the marsh pushed 
outward over the newly formed land.' 

In this way, step by step, we separate out the three elements which take 
part in the formation of the coal-bearing sediments: these are (I) the 
marine beds — as a rule limestones, sometimes shales; the limestones of 

^ J. Beete Jukes, The South Staffordshire Goal-field, 2nd ed., 8yo, London, 1859, p. 87. 

* A. de Lapparent, Traits de g^ologie, 2^ ^d., 8to, Paris, 1885, p. 841. 

' E. B. Andrews, Some Conclusions, Theoretical and Pnu^tical, Rep. €^1. Suit. 
Ohio, I, 8yo, Columbus, 1873, pp. 345-364; and by the same, On the Parallelism of 
Coal Seams, Am. Joum. Sci., 1874, 3rd ser., YIII, pp. 56-59; J. S. Newbenj, On 
the Parallelism of Coal^ Seams, op. dt., 1874, 3rd ser., YII, pp. 367-369 ; J. J. Stevenson, 
The Upper Coal Measures west of the Alleghany Mountains, Ann. Lye. Nat. Hist. 
N. York, 1873, X, pp. 226-252 ; and Note on the Coals of the Kanawha YaUey, West 
Yirginia, tom. cit., pp. 271-277, map; by the same, On the alleged Parallelism of 
Coal Beds, Proc. Am. Phil. Soc. Philadelphia, 1874, XIY, pp. 283-295. Long ago the 
same question was also eagerly discussed by distinguished American geologists; 
Rogers insisted on the splitting of the coal seams, Lesquereux on their formation 
in place, a theory which is evidently very inconsistent with the splitting. 



CH.V] PARALIC COAL-MEASUEES 247 

moderate thickness persistent over wide areas, the result of slow continuous 
formation; (2) the beds of sandstones or shales — swelling out from paper- 
thin partings into great wedge-like masses^ sometimes showing signs of 
a rapid deposition, such as oblique lamination or pockets filled with sand ; 
(8) the eoal beds — some of them thick deposits formed off the shore, split- 
tang up towards the sea through the thickening out of the unproductive 
meaflores in the manner just described. 

A oorreei eoneeptian of the facta long ago led Naumann to distinguish 
UnMiie and paralie measures, namely those which indicate marshy surround- 
ings and those which point to marine conditions. Naumann even at that 
early time recognized that in the paralie regions numerous coal beds occur, 
but that they are as a rule of trifling thickness \ In like manner Qiimbel 
distinguishes autochthonous measures, L e. those which have been formed in 
place ; and aUockthonous measures, those which have been produced by 
drifted plant remains deposited in water K 

We must admit, as is shown by the observations on the splitting of 
eoal seams, that a thick limnic bed may subdivide into a number of paralie 
seams, the latter being certainly allochthonous. The widely extended beds 
which occur in frequent alternation with marine sediments in the central 
part of the United States belong to the paralie allochthonous formationa 
It must be left to later investigations to decide whether the abundant 
rainfall which perhaps characterized this period may not have brought 
about the accumulation of vegetable detritus in shallow ill-defined lagoons, 
and so produced the limnic beds (the view advocated by Qrand'Euiy and 
shared by Saporta ^), or whether these beds may not after all be autoch- 
thonous, formed from plants the roots of which still remain in place. 

It is certain that during the Coal period the moist land was thickly 
overgrown by a peculiar vegetation. The sudd of the Nile and the floating 
mattresses of plants of the Mexican lakes bear witness to the luxuriant 
development which may be attained by vegetation in a warm climate and 
in fresh water. When Cameron was searching in 1874 for the outlet 
of lake Tanganyika in the direction of the Lukuga, he travelled some miles 
down the river until the masses of floating plants made further progress 
impossible. The lake rose, and in 1876 Stanley received from all sides 
accounts of the occurrence. Palms, which in 1871 stood on the shore 
at Ujiji, were now standing out in the lake 100 feet from shore. ' The 
whole country will be inundated and nothing will be left except the tops 
of the great mountains,' said one of the native chiefs. Stanley embarked 
on the Lualaba, and describes the gradual hemming in of the river by the 

1 C. F. Nanmann, Lehrbuch der Geognosie, 2. Aufl., Leipzig, 1862, II, p. 584. 
' C. W. Ton Gdmbel, Beitr&gezurEenntniBBderTezturyerhftltninederMiiieralkohleii; 
Sits. k. bajr. AkacL MOnchen, 1883, pp. 111-216^ pi. 
* G. de Saporta, Bull. Soc. g^ol de Vr., 1887, 3« s^r., Y, p. 883. 



248 PALAEOZOIC SEAS [pABrm 

floating masses of plants, which approached from right and left, until they 
finally closed up round tranquil standing water. The depth of the water at 
the wall:like barrier, which was covered with high reeds, was seven to 
eleven feet. Pressing forward about twenty yards into the reeds he found 
impassable mud, black as pitch, and seething with animal life. He foretold 
the bursting of the dam. 'The accumulated waters of over a hundred 
rivers will sweep through the ancient gap with the force of a cataclysm, 
bearing away on the flood all the deposits of organic d^ris at present 
in the Lukuga creek.' This rupture occurred no later than 1878. Shortly 
after, the missionary Hore descended the current of the Lukuga, now set 
free, which was then extremely rapid. When Lenz visited Tanganyika in 
1886 the lake had fallen fifteen feet. Relict strand-lines were seen between 
Ujiji and the lake, and in many other places \ 

On the leeward side of the lesser Antilles, Agassiz met with leaves, 
pieces of sugar-cane, and fragments of decomposing plants, ten to fifteen 
miles from the land, and at a depth of more than 1,000 fathoms ; and near 
some of the Polynesian islands the Challenger sometimes dredged a leaf or 
a branch from more than 1,400 fathoms. Notwithstanding these instances, 
the wide dispersal of decomposed plant remains in sufficient quantity to 
form seams of coal still remains a marvellous phenomenon. 

As regards the questions under consideration, it is clear from what has 
been said that we are scarcely justified in assuming that eveiy marine 
intercalation which occurs between paralic measures represents an oscilla- 
tion of the strand. The splitting of the limnic coal definitely excludes 
such a supposition. Only where beds certainly autochthonous alternate 
with marine beds, or where littoral inundation manifests itself over a wide 
area, can oscillation be regarded as proved. 

When the paralic coal seams alternate frequently with marine beds, 
as for instance in Illinois, and split up still further till they pass into 
unworkable beds of vegetable detritus, then they approach more and more 
closely to the character of those shaly partings which, repeated often 
hundreds of times, divide our limestones into definite beds. Since these 
paralic formations are known, both beneath the great mass of the stratified 
Carboniferous limestone as well as above it, the question arises whether 
the clayey partings between the bedding planes which determine the 
stratification of the Carboniferous limestone itself may not have a similar 
origin. 

The processes which we have been tracing in their effects took place 
very quietly, and were prolonged over periods of extraordinary duration. 
Still it is incontestable that in the middle of the Coal-measure period 
a number of great folded ranges were both elevated and worn dovm by 

^ H. M. Stanley, Through the Dark Continent, London, 1878; 0. Lenz, Brief, 
Mitth. k. k. Geogr. Gea., 1887, XXX, p. 98. 



CH. V] 



THE PEKMIAN SYSTEM 



249 



denndation, and the younger measures passed in transgression over the 
abraded surfaces of the folds into which the earlier measures had been 
thrown. This is the case not only in Europe, but in all likelihood in Nova 
Sootia also \ In China, the folds of the Tsin-ling-shan involve the Car- 
boniferous limestone, but on their denuded surface Richthofen encountered, 
at a great height, transgressive Coal-measures. 

5« The Permian eydem. It is difficult to form any conception of the 
protracted period which would be necessary to accomplish the foldings and 
tiie erosions* of the Carboniferous period. The magnitude of the erosion 
appears to its full extent when we take into account the Permian forma- 
tion, during the deposition of which the sea again returned over regions it 
had previously abandoned. 

Let us first consider Bohemia 

In the region of lower Silesia and Bohemia, according to SchUtze's 
investigations, the complete series of the Coal-measures is represented, 
from the Culm to the Rothliegende. Five successive floras may be 
distinguished, united it is true by only a few species in common. The 
Waldenburg group at the base, approximately on the horizon of the 
marine intercalations of upper Silesia, contains fishes, Estheria, and possibly 
Modiola. But we can hardly regard this as representing the Gannister 
beds, elsewhere of such wide distribution. A good deal of conglomerate is 
int^x»Iated with the beds K 

In central Bohemia, and thence beyond Pilsen and Merschan, the 
Carboniferous lies horizontal and unconformable on the edges of the 
Silurian system. The latter, together with the associated subdivisions of 
the Devonian, was already faulted down, and doubtless completely sepa- 
rated from the other Silurian and Devonian ranges, at the time when the 
Coal-measures were formed over its surface. Here, as we have seen, some 
basal beds of the Coal-measures occur, which sometimes adapt themselves 
completely to the slight inequalities of the older rocks lying beneath them ; 
these are followed by a great thickness of unproductive clastic rock, and 
above this come the uppermost measures, perhaps of lower Permian age, 
and the sandstone of the Rothliegende. 

The Carboniferous mantle of central Bohemia dips to the north-west 
under the Rothliegende, and its termination is not visible. To the south- 
east, however, it is broken up by erosion into larger or smaller outliers, 
and over considerable areas has been entirely destroyed. 

This limnic transgression of the Carboniferous, however, extended 



^ J. W. BawsoD, On the Upper Goal-Formation of Eastern Nova Scotia and Prince 
£dward Island in its relation to the Permian; Quart. Joum. GeoL Soc, 1874, XXX, 
pp. 209-219. 

* A. Schfltze, G^ognostische Darstellang des nieder8chle8.-b0hmi8chen Steinkohlen- 
beckens ; Abb. z. Specialkarte v. Preuss., 1882, III, p. 19. 



260 PALAEOZOIC SEAS [PABTin 

farther and farther over Bohemia, first filling ap the hollows and then 
gradually levelling down the land itself. Hence we see isolated patches of 
the appermost Coal-measures preserved in the marginal fractures towards 
Moravia and Bavaria, while they have been entirely swept away from the 
higher parts of the Bohemian mass. The transgression continued during 
the formation of the last beds of coal and into the Bothli^^ende. Henes 
a little isolated patch of Permian ooal-measures still lies near Budweis in 
the midst of the Archaean region, and the remains of the denuded ooveimg 
of RothUegende crop out along the marginal fractures to the south-west as 
far as Begensburg, and to ihe south-east as far as Zobing, near Krems. 

The sea, however, did not follow these transgressions. They recall in 
a striking manner the ridging upwards of the borders of the Pittsboig 
marsh during the positive phase of the sea as conceived by Stevenson. To 
me also it is difiicult to imagine the progress of these limnic transgressiaDS 
without a simultaneous rise of the strand. They are everywhere repre* 
sented in central Europe, first by the upper Coal-measures of the Carboni- 
ferous (Badowenz or upper Ottweil group), then by the Permian measnres, 
and finally by the widely distributed Bothliegende sandstone and the 
accompanying conglomerates. This sandstone also extends far over the 
Bussian plain, here and there accompanied by gypsum or rock-s&It 
Finally the positive influence became so preponderant that the sea re- 
appeared over a considerable part of central Europe. This was the period 
of the Zechstein with its impoverished marine fauna. From Buasia this 
formation extends across the whole of north Germany; dolomitic pro- 
longations, the edge of the lens as it were, reach England. Everywhere 
it follows the Bothliegende, nowhere does it pass beyond, readiing at 
most the extreme outer border of that series; thus it is well known in 
Silesia, but not on the Bohemian mass. 

The Zechstein marks a recurrence peculiar to central Europe and to 
a part of northern Europe. It does not penetrate into the Alps. On the 
other hand, it appears far away, in Kansas, represented by beds overlying 
the Fusulina limestone and containing a fauna comparable with that of 
the European ZechsteiiL 

We have now reached one of the most remarkable phases of the earth's 
history ; and, putting detail entirely aside, we will carry our survey beyond 
the confines of Europe and the United States, and endeavour to obtain 
a general view of the whole course of events. 

Let us once more return to the base of the Carboniferous system, or 
the Calciferous limestone of Scotland, the equivalents of which we have 
already recognized on the other side of the Atlantic Ocean. The stratified 
series of Fife, with its still-rooted Stigmaria standing between the marine 
intercalations, affords clearer proof than any observations hitherto made 
of repeated oscillations with a positive excess, a phenomenon not to be 



V] 



CAEBONIPEEOUS OSCILLATIONS 



261 



ecmfonnded with the alternation of paralic beds. In Illinois and the 
south-east of Russia the oontemporaneoos formations were purely marine. 
Poriiaps the whole series of sandstone and coal beds which extends from 
Banks land over the north coast of Melville sound to Baffin bay must 
alao be referred to the same origin. 

The positive movement now conducts us to the most important pelagic 
member of the period, the Carboniferous limestone, which reveals, not only 
by its calcareous composition and its rich marine fauna, but above all by 
its mighty transgression, the far-reaching changes in the geography of the 
coast It rests normally on the older Palaeozoic series from the far North 
to BrazU and Australia, but at the same time exceeds their limits. In 
Ireland, England, Scotland, and Spitzbergen it lies as a purely marine 
limestone on the Old Bed sandstone, the only characteristic fossils of which 
are Ganoids and terrestrial plants; it extends at the same time through 
the Parry islands into the highest known latitudea In central and 
southern Greenland it has not been observed. It proceeds through a 
large part of the great ranges of Asia, and in east China it rests on the 
Devonian and Silurian of the Tsin-ling-shan, overlaps them north of the 
Wei-Ho, and stretches along the sea from the north of Shensi to 
Shan-tung and liao-tung, resting everywhere within this ancient table- 
land in apparent concordance upon the Cambrian beda In like manner 
in the east part of the United States it rests on Devonian, extending to 
the west and south-west. In Dakota and the Bocky mountains of Colorado 
it follows in deceptive conformity upon Cambrian beds as in north China; 
at the bottom of the Grand Cation a great plane of erosion separates it 
from the Cambrian or still older beds, but in the table-land of Texas, 
Cambrian rock reappears, immediately overlain, and again in apparent 
concordance, by the Carboniferous limestone. In California, so far as we 
know at present, the Carboniferous limestone is the sole representative of 
the entire Palaeozoic series. 

This great transgression is succeeded, so far as we can judge from the facts^ 
by a very marked and extensive negative phase. The flora of the lowest 
horizons of the resulting Coal-measures^ i.e. the Culm flora, is known in 
the Arctic regions, in Europe, at many places in Siberia, and as far as 
Australia, where it is represented by several characteristic species. Thick 
masses of clastic sediments overlie the Carboniferous limestone in the 
United States, in the east of Canada, in central Europe, and in China ; in 
central Europe the marine Gannister beds are intercalated with the lower 
Coal-measures, and after a remarkable sinking of the strand the signs of 
renewed ascent are revealed in the thickness of the sediments and the 
continuance of the limnic transgression. 

This ascent finally led to a renewed appearance of the sea. It is the 
upper Carboniferous sea of the Fusulina limestone. The plant remains 



262 PALAEOZOIC SEAS [pAKrni 

determined by Stur — which lie in the shales between the beds of the 
Fus'ulina limestone on the Kronalp in Carinthia — leave no room for doubt 
as to the contemporaneity of this marine limestone with the higher Bado- 
wenz or upper Ottweiler Coal-measures, that is with the uppermost 
flora of the coal basin of lower Silesia and Bohemia. This new marine 
deposit, however, does not extend nearly so far to the north as the Car- 
boniferous limestone. In nUnois it alternates with the Coal-measures, 
thins off towards Ohio and north Virginia, and does not reach Pennsyl- 
vania or Canada. In Europe it is found in north Spain; Meuniw 
encountered it on the Morvan ; in the southern Alps it alternates with 
plant-bearing beds; then it reappears in the south of Russia; Neumayr 
has recognized it in the north-west of Asia Minor ; Teller in the island of 
Chios. Everywhere the sea appears to have come from the south. Simi- 
larly, all the specimens of Fusulina limestone so far described by Schwager 
from Bichthofen's collections, and the rich upper Carboniferous fauna of 
Loping in the province of Eiang-si, lat. 27* 52' N., studied by Eayser, 
belong to districts which lie south of the Tsin-ling-shan. 

Now came the extension of the BotUiegende over Bussia and central 
Europe, and even into some parts of the southern Alps ; then, the positive 
movement still continuing, the Zechstein followed, extending from Buasia 
through north Oermany into England. It was a shallow sea; salt and 
gypsum crystallized out from it. During the period of this transfgression, 
which came from the north, and perhaps also during the Bothliegende, a 
new fauna was introduced from the south, which has not yet received 
a particular name. It includes Carboniferous and Permian species and 
some new types, and is known to extend from the Salt range of India 
through the great ranges of Asia, Afghanistan, Djoulfa, in the valley of 
the Araxes, and past Artinsk in the Ural to Sicily. Its most northerly 
representative is perhaps the fauna of the Bellerophon beds in the south- 
eastern Alps; these beds, however, like the Zechstein, rest upon ihe 
Bothliegende \ 

While submergence and emergence thus alternate over regions of such 
vast extent, the outlines of a great continent become disclosed to us, and 
from the closing days of the Carboniferous this remains for a long period 
one of the most prominent features of the face of the earth. On a previous 
page (I, p. 387) it has been described as the fractured Indian continent^ 
or Oondwdna land. At the present day it is broken up by the Indian 
Ocean, and comprises the peninsula of India, Australia, and a large part 
of Africa. 

^ E. yon MojsisoyicB, Ueber das Yorkommen einer mnihmasslich TOitriadiBchen 
Cephalopoden-Fauna in Sicilien, Verb. k. k. geol. Reichs., 1882, p. 31; G. Stache, 
Zur Fauna der Bellorophonkalke Sadtirols, Jahrb. k. k. geoL Beichs., 1877, XXVII, 
pp. 271-318; and 1878, XXVIII, pp. 93-168, pi. 



CH. v] ANCIENT GLACIATION 253 

The investigationfl of recent years, and the discoveries made by Warth 
in the Salt range, have helped to complete our knowledge of this continent. 
In the Salt range erratic blocks transported by ice occur at a horizon 
\rhich, according to Waagen, must be correlated with the highest sub- 
divisions of the Carboniferous system \ Th^ are the same accumulations 
which constitute the Talchir conglomerate, or the lowest stage of the 
plant-bearing G(ondw&na series, in which Blanford, Oldham, and Fedden 
have recognized the influence of ice (I, p. 404); they also form, as the 
Dwyka or Ecca conglomerate, the foundations of the Earoo formation in 
Africa, and were long ago declared by Sutherland to be of glacial origin 
(I, p. 389), 

The same glacial beds also make their appearance in east Australia, as 
the Bacchus marsh or Stony creek beds. There they alternate with marine 
beds which are referred to the Carbonif eroua 

They are immediately succeeded in India, Africa, and Australia by the 
long series of plant-bearing beds which ascend high into the Mesozoic 
group. For a long time the whole region remained above the sea. Then, 
during the Mesozoic aera, it gave way, breaking up piece by piece, the 
process probably continuing down to a comparatively late period, and 
laige parts of it have not even yet been reached by the sea. 

According to Waagen the glacial region extended from lat 40"" S. to 
lat. 35"" N. and from long. 18'' £. to long. ISd"" E. Some typical plants of 
the Culm flora are known in Australia, but the upper floras of the Carboni- 
ferous of Europe and North America have not been discovered on this 
continent. Over the whole of Qondwlma land an independent series of 
floras makes its appearance under the influence of a severe climate. 

In Europe erratic blocks of wholly enigmatical origin have been met 
with in the lower horizons of the Coal-measures, e. g. in proximity to 
a marine intercalation near Ostrau. According to Ramsay and Geikie 
signs of glacial action are to be met with in the Bothliegende of England. 

Opinions as to what are the precise equivalents in Europe of the glacial 
conglomerates of India are not yet in absolute agreement. According to 
Waagen, as already mentioned, they should be placed on the horizon of the 
highest divisions of the Carboniferous ; in Australia, glacial conditions would 
seem to recur in the Permian period. 

The Carboniferous limestone of the far north contains a number of 
species also found in the Zechstein ; its relations to the Zechstein do not 
appear to me to be as yet completely explained. 

6. Survey of the Palaeozoic seas. The question whether the changes in 
the distribution of the seas can be explained by a secular rise and fall of 
the continents may receive some illumination from the history of the 
Palaeozoic seas. 

^ W. Waagen, Die carbone Eineit, torn, cit, 1887, pp. 148*192. 



254 PALAEOZOIC SEAS [pabtih 

We have recognized the existence of two continents, of which fragments 
only are visible at the present day. The first occupied the position of the 
north Atlantic Ocean, as is indicated by the nature and distribution of 
the Palaeozoic sediments in Europe and America ; Greenland is a remnant 
of it. This ancient continent is the Atlantis. 

The second continent, first clearly discernible towards the close of the 
Carboniferous period, is now represented by three fragments, Africa, India 
and Australia. As Greenland on the one hand, so the Indian peninsula on 
the other projects into the Ocean which covers the subsided table-lands. 
This continent is Oondwdria Icmd. 

The destruction of the two continents was accomplished later, piece by 
piece, and we can follow the process through many stages. It must have 
produced many negative movements and thereby have led to the abandon- 
ment of other regions by the sea. Further, our stratigraphical inquiries 
have so far advanced that it has become possible to recognize a number of 
positive and negative displacements which affected very extensive areas in 
a uniform manner. 

At the clos6 of the Silurian period the sea became shallower over a wide 
region, from Illinois and Iowa to Wisconsin and New York, in England, the 
Bfdtic provinces, and as far as the Dniestr. Then the Old Red sandstone 
advanced over a great part of the Atlantis. Evidence of its encroachment 
may be observed in Scotland, England, the gulf of Finland, and as far as 
the gulf of Onega and Spitzbergen, perhaps as far as the south of Green- 
land ; and its equivalents, retaining the same .characters, appear in New 
Brunswick. Towards the middle of the Devonian, the marine transgression 
extended over these continental or at most sublittoral deposits, and attained 
its maximimi. The limestone and dolomite deposits of the middle Devonian 
extend over Orel and Voronezh to Livonia and Courland, and at the same 
time the transgressive middle Devonian appears to form the western glint 
of the Canadian shield from the Clear- water to the Arctic Ocean. In the 
oil-bearing beds on the shores of the Uchta, in the region of the Petchora^ 
we recognize the oil-bearing shales of the banks of the Athabasca, which 
may be traced in North America from the valley of that river as far as the 
Gaspd Then the shore again receded for a great distance, and the Carboni* 
f erous period began. 

From the limnic or littoral formations, which in some localities seem to 
show signs of frequent secondary oscillations with a positive preponderance, 
we pass to the Carboniferous limestone. Simultaneously with the appear- 
ance of pelagic characters, the transgression is found to advance beyond the 
limits of earlier deposits ; it occurs in regions the most remote from each 
other, as in China and Texas. 

Once more the strand receded over a wide area, and then again advanced, 
evidently from the south, but it did not proceed nearly so far as during the 



CH.V] SOUTHERN SEAS 266 

Carbonif erous limestone periocL This is the sea of the upper Carboniferous 
or FuBulina limestone. 

Tet another marine fauna made its way from Asia across the Araxes, 
past Artinsk, to Sicily ; whether or not it is represented in the south-eastern 
Alps is somewhat uncertain* It did not attain to the extension of the 
preceding stage. It is not known in the United States. Simultaneously 
with this fauna, the sea came down from the north and spread out the 
Zechstein over northern and central Europe. 

Much- no doubt still remains obscure. The Palaeozoic deposits of the 
Sahara, Brazil, and other very extensive regions, can hardly be taken into 
account in comparisons of this kind, and in the present state of our know- 
ledge it is not possible to oppose in the Palaeozoic aera a compensatory 
n^ative region against a given positive region. Nevertheless we can 
aheady recognize three faunas which are foreign to the north of Europe, 
namely^ the Hercynian stage, the Fusulina limestone, and the stage of 
Djonl&k On the other hand the typical Zechstein is not known in the 
southern regions. 

Positive and negative movements alternate simultaneously over regions 
of such vast extent, that they cannot be explained by a bulghig or a sagging 
of the lithosphere on however great & scale. During the Carboniferous and 
at other periods also, considerable foldings of the lithosphere occurred, but 
they have nothing in common with this phenomenon of general inundation 
and emeigence. The Armorican and Yariscan folds arose in complete 
independence of the transgressions, by the progress of which these folds 
themselves were planed down and covered up. 



256 



CHAPTER ri 

MESOZOIC SEAS 

The seas of the Trias. Pontive movements in the Rhaetic epoch. Continuation of 
positive movements daring the Jurassic period. Negative phase in central Europe and 
commencement of the Cretaceous period. Later transgressions and intermixture of the 
Cretaceous faunas. The Cenomanian transgression. General survej of the 



1. Seas of the Triae. In the east of the United States the marine 
Mesozoic series is absent nntil we reach the middle Cretaceous. The 
Rhaetic and Lias lignites of the Appalachians clearly show that during 
their formation this region was not covered by the sea. Towards the west, 
however, first, marine upper Jurassic is met with in the Black mountains 
of Dakota, then the marine Trias makes its appearance in the western part 
of the Basin ranges, and last the Neocomian in California. 

In Brazil likewise, marine sediments are completely absent from the 
beginning of the Mesozoic aera up to the middle of the Cretaceous. Even 
far to the west, in Jujuy and Salta, the northern provinces of the Argentine 
republic, we still meet with Rhaetic plant-bearing beds. Still further to 
the west, however, in the Andes of Chili and Peru, we encounter various 
marine members of the Trias, the Jurassic and the lower Cretaceoua 

In eastern Australia there is a similar absence of marine sediments 
from the beginning of the Trias upwards, but the gap extends in this case 
only as far as the lower Cretaceous ; the whole of this interval is repre- 
sented by a few plant-bearing sediments only, such as the Jerusalem beds 
in Tasmania and the Clarence river beds in New South Wales and Queens- 
land. In New Zealand, on the other hand, the marine Trias is present as 
well as marine members of the Jurassic. Here the series does not appear 
to be quite complete, and some plant-bearing sediments again alternate 
with marine beds. 

In the north-east of China the marine series terminates in general with 
the Carboniferous ; not even the marine deposits of the middle Cretaceous, 
elsewhere so widely distributed, have hitherto been encountered there. 
But the marine Trias, together with some members of the marine Jurassic, 
is known in the moimtain-chains of Japan. In Tezo and Saghalin only 
the middle Cretaceous has so far been observed. 

In the arc of the Aleutian islands, Trias, Jurassic, and Cretaceous are 
represented by marine deposits. 



CH.TI] THE OCEANS OP DIFFERENT AGE 

We may thus faiiiy conclade that in proportion aa we apptxiacli 
Pacific Ocean that important part of the Heeozoic aeries which is ( 
Uian the middle Cretaceona becomes more complete. 

The Atlantic border difiers completely. With the exception of the 
r^t(Hi8 consbmcted on the Pacific type, the Antilles and the strail 
Gibraltar, perhape also the oatnmnera of the chains as far as the 1 
Draa and some parts of Europe, marine Mesozoic sediments older thai 
middle Oetaceoos do not appear at all on the Atlantic coasts. This ii 
caae from cape Horn to beyond the mouth of the Orinoco, from Fl( 
to the Arctic Ocean and cape Farewell, and from the Wadi Draa tc 
cape of Good Hope. 

The Indian Ocean presents both types. On the shores of Ara 
where the Pacific structtire is the rule, the marine Trias is known to ( 
but it is not found anywhere else along the coast of this Ocean. ' 
here also a great gap occurs in the lower half of the Mesozoic series 
it does not extend as far as the middle Cretaceous, and indeed no fni 
than the middle Jurassic. 

From this it follows that tiie eadating Ocecms an of different age. 

The distribution of Petudonumatie Oehotiea, one of the most chi 
teristic species of the Trias, is of particular interest; discovered, ai 
have already mentioned, by Middendorf in Mamgi bay (sea of Okhots 
is recorded la Teller's tables as having been found subsequently at a j 
number of other localities ; at several places in Japan, particularly ii 
bay of Sendai; on the island of Hugon; in New Caledonia; at se 
places in the Alps of New Zealand; as a closely allied variety, or 
rio Utcabamba, between Chachapoyas and Cuelap, Peru ; in Fli 
county, California, Humboldt mountain, Nevada, uid at several i 
places in these chains ; near fort Rupert, Vancouver ; in Moresby is 
Queen Cbu'lotte group ; on the Peace river, on the east slope of the R 
mountains (lat. 50° to 57* N.) ; at cape Nunakalkhak, at the entrani 
the bay of Povalak, Alaska ; finally near Verkhojansk on the Jama in 
Siberia (hit. 68° N.)'. 

Thus the marine Trias engirdles the Pacific Ocean; it is conti 
across Siberia from the mouth of the Olenek to Spitzbergen, but it 
not reach the Atlantic region by this route. 

Mention has been made (I, p. 681) of the marine Trias of An 
Its thick aeries of sediments help to form the great ranges of Asia 
their connexion with the Trias of the Alps is still little known. C 
bach's widely extended investigations in Afghan Turkestan show 
Halobia Lom/mdi of the upper Alpine Trias recurs near Chahil, north- 

■ F. Teller, Die PelecTpoden-Faona ton Wercbojanak in OitBibiriea, in E. Hojai 
TonHojgrir, ArktiiclieTriaafatiiien', H^m. Acad. Imp. Sci. BBint-P^tenb.,1886, X} 
No. 6, p. 108 et Mq. 



268 MESOZOIO SEAS [ 

of Saighan in Turkestan^. But it appears there associated with plant- 
bearing beds, which represent part of the Qondw&na series of India, and 
we have seen that a lower part of this series enters into the stmctnre of 
the great ranges near DarjiUng (I, p. 577). 

Further to the west information is wanting. In Russian Turkestan 
the Triaa is absents In the Caucasus only plant-bearing beds are known 
below the lias, to which they are supposed still to belong. MojaLsovics, 
who has rendered such great service in determining the subdivision of the 
Trias, believes he has found indications of the lower Alpine Trias above 
the Permo-Carboniferous sediments of Djoulfa in Armenia; but it was 
the collections sent by Russian geologists from the mountain of Bogdo on the 
lower Volga, which first enabled him to prove definitely the presence of the 
Alpine Trias in this region. In the north-west of Asia Minor, on the lower 
Danube, in the Balkans, in the Carpathians> and in the Mittelgebirge of 
Hungary, this formation appears with increasing frequency until we reach 
the region of the eastern Alps. From here the sediments of this group 
extend to the Orisons, the southern Alps and Bosnia^ and proceed through 
the Apennines to Sicily; they are found, though not with the same 
characters, in the northern slopes of the Swiss Alps and cross the Balearic 
isles to the Ebro. 

These facts reveal to us a sea which extended from the interior of 
Asia over the south of Europe. It is this sea which the study of the 
distribution of the Jurassic sediments disclosed to Neumayr, and it has 
been named by him the 'Central Mediterranean^.' 

We have seen from three examples, namely from the Hercynian stage 
of the lower Devonian, from the Fusulina limestone of the upper Car- 
boniferous, and from the Permo-Carboniferous deposits of Djoulfa, that 
calcareous marine formations oecur in the south and south-east of Europe; 
while clastic sediments and the deposits formed in shallow water prevail 
towards the middle or the north-west The same phenomenon presents 
itself again in the Triaa Even in the regions where the system firat 
received this name (central and northern Germany, together with England 
and a large part of Fiance, including the Jura and the D^partement du 
Var), it presents itself with a local and abnormal f acies : clastic and sub- 
littoral, lacustrine, and saliferous deposits assume far greater prominence 
here than elsewhere, and its development in these re^ons aa the triad 
group, Bunter sandstone^ Muschelkalk, and Keuper, affords a striking 
example of a sedimentary cycle, with a lenticular mass of limestone as 
the middle term. The Muschelkalk indeed is absent towards the north- 

^ C. L.Griesbach, Field Notes : No. 5, to accompany a Geological Sketch-map of Afghsni- 
Bt4ii and Korth-eattem Khorassan ; Bee. Qeol. Surv. India, 1887, XX, p. 97 et aeq. 

' M. Neamajr, Die geographische Verbroitong der Jaxaformation ; DenVachr. k. Akad. 
Win. Wien, 1885, L, pp. 57-86, maps. 



CH.VI] TRIAS IN SOUTH TYROL 259 

west, namely in the wefitezn part of the basin of Paris, and in England ; 
there the Kenper rests directly upon the Bnnter sandstone, often only with 
difficolfy to be distingnished from it. 

It was impossible to arrive, at a correct interpretation of the most 
important of these regions, so long as the horsts of the Rhine, the Black 
Forefit, and the Yosges were regarded as parts of an ancient coast. Benecke, 
as early as 1877, had liberated himself from this mistaken theory, and was 
then led to formulate general views in respect to the west of central 
Europe \ But if we follow in detail the descriptions of Benecke, Sand- 
beiger, and other Gterman investigators, we perceive the danger of a too 
simplified and too generalized conception of the Trias as representing 
a regular cycle. The Bunter sandstone is indeed a transgressive deposit 
marking a positive period; it rests on rocks of very different age. But 
the beds of various nature which are intercalated in this series, some- 
times containing marine shells, at others terrestrial plants, and composed 
in some cases of dolomitic limestone ; or again, the variegated gypsif erous 
days of the middle Muschelkalk, the various salt-bearing zones, and the 
repeated interbedding of dolomite with the Eeuper, which includes here and 
there, as far as north (Germany, isolated marine moUusca, and near WUrz-> 
burg even Myophoria Baibliava, so widely distributed in the eastern 
Alps — ^all these are only a few of the many anomalies, probably true 
recurrences, which interrupt the cycle. At the same time, with the progress 
of events, the sediments continued to increase in thickness, until finally, at 
the dose of the Keuper, the littoral sediments of the Rhaetic, which forms 
the next stage, extended beyond the limits of the Trias on to the older 
rocks, indicating a higher level of the coast-line than had at any time 
existed during that period. 

The calcareous deposits of the extra-Alpine Trias are always, wherever 
they occur, intercalated between the two clastic members, the Bunter sand- 
stone and the Keuper. Their distribution corresponds in a striking manner 
with that of the Zechstein, but is not so extensive. Here we have a cycle 
accomplished in oscillations. 

Contemporaneously, sediments of quite a different nature were deposited 
in the eastern Alps. They are almost exclusively of pelagic origin and 
distinguished by a rich fauna ; limestone and dolomite attain a considerable 
thickness in this region. 

A very singular development characterizes the Trias in south Tyrol. 

F. von Richthofen observed that at certain horizons of the Trias in this 
region vast masses of limestone and dolomite protrude, forming mountains 
with steep slopes, and that between these masses and on their flanks dastio 
sediments and volcanic tuffs have been deposited. He regarded them as 

* E. W. Benecke, Ueber die Trias in ElsaM-Lothringen und Luxemburg ; Abb. s. geoL 
Speeialkarte ▼. ElsaM-Lothringen, 1877, 1, Heft 4. 

S 2 



^ 
>" 



260 MESOZOIC SEAS [PABrm 

coral reefs. Qiimbel raised the objection that corals are rare in these 
limestones, and that for the most part they are stratified The facts of the 
case are clearly set forth in the descriptions given by E. von Mojsisovics 
and his fellow workers, H. Hoemes and C. Doelter \ Mountainous masses 
of limestone and dolomite are buried up in the clastic sediments of the 
Trias. The form of these masses does not correspond with that of the 
mountains, except in those rare cases where the outer slopes of the reef 
have been exposed by denudation, and persist as the existing slopes of the 
mountain. Most of the masses are stratified, but some not. Mojsisovica 
thinks the unstratified masses are coral reefis. On the outer slope of such 
masses ancient accumulations of d Aris with * overcast-bedding ' are some- 
times to be seen. That the expression 'reef is justified can hardly be 
doubted. In the lower horizons we may observe between the Adige and 
the valley of the Sexten two regions of reefs, a western and an eastern; 
in the intervening district the reefs are whoUy missing. At a later period, 
when the Saint Cassian beds had been deposited, these two regions of reefs 
were united in a band which runs through Caprile, Reve di Cadore, and 
Auronzo, so that the area of deposition of the clastic sediments was 
diminished. We may add that at a still higher horizon of the Plattenkalk 
the stratified beds extended over the whole region. 

In the northern Alps massive or stratified limestones and dolomitee 
have been found in more or less continuous sheets. We will restrict our 
attention to the upper beds, which are so intimately associated with the 
Bhaetic stage, 

2. Positive movemeTUs in the Bhaetic period. Standing near the 
Konigssee near Berchtesgaden, we may readily distingmsh in the white 
wall of cli£b above Sankt Bartholomae two weU-marked subdivisions: the 
rock below is grey, unstratified, splintery, and weathers into numerous 
sharp cones ; that above is divided by regular bedding-planes and crowns 
all the higher summits of the cliffs. But if, south of tiie lake, we ascend 
the height of the Steinemes Meer, the eye sweeps over a panorama of 
mountains which, as far up as the peak of the Watzmann, are traversed 
by the iaame regular bedding-planes. This member of the Alpine limestone 
immediately underlying the Bhaetic stage has received the characteristic 
name of PlcMenkcUk. It has been traced from the Yorarlbeig to the 
eastern extremity of the Alps near Vienna, as well as through the whole 
limestone region of the southern Alps, and, maintaining the same characters, 
in the faulted-in band of the limestone which traverses Carinthia. We 
may also include with it the stratified limestone masses of the Dackstein 

^ F. yon Richthofen, Qeognostische Beschreibung von Predano, St. Oassian and der 
Seisaer-Alpe, 4to, Gotha, 1860 ; C. W. Gtimbel, Das Mendel- and Schlemgebirge, Sibber, k. 
bayr. Akad., 1878, p. 71 et seq. ; E. MojnaoTics von Mojev&r, Die DolomitrifPe von Sfld- 
Tyrol and Venetien, 8vo, Wien, 1879. 



OH. VI] KHAETIC POSITIVE MOVEMENTS 261 

tnimntaina, and r^ard Flsttenkalk and Dachsteinkalk as Bynonymooa 
teiXQB. The pale limestones, however, which overlie the richly foasilifennis 
beds of the Bhaetic stage, and in the weetflm port of the northern Alps 
are nsaally spoken of aa Dachsteinkalk or npper Dachateinkalk in opposi- 
tioii to Plattenkallc, are only a recurrence of the beds of the Plattenkalk at 
a higher stage. 

That the shells of foraminifera form a large part of the substance of 
these limestones was known long ago to Schafhantl and Qdmbel ; in 1863, 
Peters showed by an examination of thin slices that the hmeetone of the 
Daehstein moontains and the Terglou abounds in micro-organiBms '. Bat 
they present many other instmctive characters; a wide and attractive 
field of investigation opens out here, as yet hardly touched, and I can there- 
ton only offer a few preliminary observations. 



Fia. 37. Vlttefivm Vm Olitrlakntr, < 

Let US approach more closely one of the colossal limestone walls. We 
may first call attention to the bright red shards which are strewn through 
some of the beds. Sometimes these fragments are angular, and look as 
though they had been broken off from a hard red bed ; sometimes tiiey are 
thinly laminated in red and yellow layers ; sometimes, again, we see lying 
in the limestone one of the great shells of Uegalodus filled with red 
material up to a level line, above this either white limestone or calcspar. 
This red material is the terra rosea of the Karst and of the emerged coral 
reefs of Oceania, the residue left after the solution of limestone, and thus 
scarcely likely to have been formed beneatii the sea. 

Next the behaviour of the coraia must be considered. In the light 
grey limestone, branches of great coral growths replaced by white calcspar 
are visible, which owing to the difficulty of a closer determination are 
generally described as Lithodendron. The radiating branches present 
a rosette-like or nodular form ; sometimes, indeed, rosettes of this kind are 

> E. F. Peten, Ueber Foiaminiferea im Docluteiskalk ; JaLrb. k. k. geol. Beichi., 
186S, XIU, pp. 29S-296. 



^T-T" "^r-J 



262 MESOZOIC SEAS [part ra 

seen seated on the shell of a great M^alodns. But the lithodendton 
corals are not only scattered through the limestone. There are some beds 
which are entirely made up of them; their branches are more or lees 
vertical, the intervals between them being filled -up with limestone. These 
are true coral limestones. At the base these beds are separated by 
a bedding-plane from the un.derlying limestone, but strange to say they 
are similarly separated above from the next succeeding bed by a continuoos 
plane of division which may frequently be followed for a great distance 
along the face of the diff: their knoU-like masses and branches do not 
extend into the next limestone bed, contrary to what, from the irregularity 
of their growth, we might naturaUy have expected. I even remember to 
have seen, several years ago, in the clifib of the Schladminger loch in the 
Dachstein mountains, two coral beds of this kind, separated by a trifling 
interval, maintaining for an indefinite distance a constant thickness. 
Whether the upper surface is a plane of erosion such as is produced by 
the sea on the borders of existing coral reefs, I am unable to say. 

The beds theTnsdvea vary greatly in character. Sometimes they aie 
grey, very splintery, and without a trace of organic remains, at others 
pale yellowish or greyish white, breaking with an even fracture, and 
occasionally rich in fossils. These beds of different nature occur in 
repeated idtemationa 

They often contain over 40 per cent, of magnesium carbonate : there 
can be no doubt that in these cases dolomite has been precipitated as such 
from the sea \ 

Beneath are beds which contain numerous minute organisms, and 
outwardly bear no resemblance to a dolomite. 

The internal structure of the beds is also veiy various. Some are 
laminated, as is clearly shown by the weathered surfaces, even in the case 
of the Gyroporella beds of the Bax Alp in lower Austria which belong to 
a lower horizon. Others contain larger or smaller fragments of limestone 
of a different origin, which are often only revealed by polishing. In the 
Dachstein mountains, on the left side of the Earls-Eisfeld, beneath the 
Schoberl, Rhynchanella ancUla is found in immense numbers. Freshly 
broken out blocks of the rock show, however, when polished, that the 
Bhynchonellas are embedded in fragments of a light grey limestone, which 
are derivative, forming foreign inclusions in this bed. The polished 
surfaces also reveal fragments of a yellowish-white limestone, containing 
other organic remains, and of an unfossiliferous grey limestone, as well 

^ Gtlmbel has always maintained tluB view; the dolomiiic intercalations in the 
Potsdam Sandstone, the Waterlime of the upper Silurian of North Ameiicfty the 
dolomitic beds of the upper Devonian of Russia, those which represent the Zechstein in 
England and those in the Glerman Keuper, all of them more or less littoral deposits, are 
so many examples of the direct deposition of dolomite. 



VI] 



THE PLATTENKALK 



263 



as parts which are coloured by red earth in bands of lighter and darker 
tints. 

A farther peculiarity is the incrudation of the included fragments. 
It not seldom happens, and an excellent instance is afforded by the Earls- 
Eisf eld rock, that each of these derivative components has been covered 
with a thick crust of carbonate of Ume, which was deposited before the 
formation of the matrix* In sections this crust shows a radiate structure. 
Occasionally adjacent fragments are enclosed in a common crust. This 
case differs completely from that in which carbonate of lime fiUs up the 
interior of shells, as for instance the brachiopods of the Vils beds, and 
replacing the rock forms in some cases the whole of the matrix as well, as 
has been described by Bothpletz ^ ; it is the incrustation of the fragments 
by sinter, followed by their cementation together, before the formation of 
the matrix and the completion of the rock. It is not easy to understand 
how this process could take place beneath the sea. 

We have next to consider the *mode of eeparaium of the htds. In some 
series successive beds are separated by sharply marked divisional planes, 
but are not distinguished by any lithological differences; in other cases 
a bed may consist of two or three layers clearly defined from each other by 
differences of colour and texture, but yet not separated by any interruption 
of continuity ^ 

It not seldom happens that a single bed consists of two or even three 
layers, which are not separated by bedding planes but by a dark, acutely 
denticulate line, not unlike a cranial suture. They have been described by 
OtLmbel and Pichler ; Bothpletz describes them under the name of * Sutures.' 
They have not been formed by pressure, as has been supposed ; if, taking 
advantage of the thin film of clay, which is seen in section as the dark 
jagged line of demarcation, we succeed in exposing the surface of separa- 
tion, we perceive that it is a true stylolith formation, i. e. that numerous 
particles of the upper deposit have sunk into the lower, forming a number 
of cones \ 

^ A. Bothpletz, Geologisch'palAoiktologiflche If onographie der Yikei^Alpen ; F^&onto- 
graphica, XXXIII, p. 66, pL xr, fig. 17. Bischof has ahown what weighty gronnde there 
are for the belief that carbonate of lime is not precipitated in the sea without the 
MButance of organisms ; Loretz has observed foraminiferae in thin slices, sarrounded in 
like manner by zones of radiate stmctare ; Zeitschr. deutsch. geol. Ges., 1878, XXX, 
p. 412, pi. xriii, fig. 11. 

* E. Saess und £. Ton Ifojsisoyics, Stndien fiber die Gliederang der Trias- and 
Jniabildnngen in den Ostliohen Alpen ; II, Die Gebirgsgrappe des Osterhomes, Jahrb. 
k. k. geoL Reichs., 1868, XYIII, pp. 167-200. 

' These satares may be seen in the polished fiices of blocks of upper Jurassic and 
lower Cretaceous limestone used in the buildings of upper Italy, but here, so far as 
I have observed them, they only separate beds of the same nature. A bed is seen on the 
Osteihom (No. 22 in the list in Jahrb. k. k. geol. Reichs., 1868, p. 172) which consists, at 
its base, of a reddish brown ; above, of a grey layer. The lower layer (a) contains many 
organic remains, including corals, and the upper grey layer (6) penetrates from above 



;s? 



,^r. 



'• •'. 



*« - T= 



264 MESOZOIO SEAS [pabtui 

We now come /to the last and most important of these stractnres, the 
partiTvgs. The bedding-planes axe usually characterized by the presence 
of an argillaceous matter, foreign to the rock. The localities where tiiiese 
intercalations or partings occur are, however, somewhat difficult to reach, 
and I cannot positively affirm that they are always present. On the 
Osterhom, south of the Wolfgang-See, thin streaks of lustrous coal formed 
of drifted stems of plants make a rare appearance here and there in the 
uppermost beds of the Plattenkalk, not far below the lowest beds of the 
Bhaetic series. The partings are formed of black bituminous shale, oon- 
taining remains of Qanoids, and ntunerous scale-like leaves and twigs of 
Araucarites aZpinus. Notwithstanding the occasional layers of coal dis- 
tributed through it, the limestone with its pale colour, its Megalodons and 
corals, presents a most striking contrast to the black shales with their 
fishes and terrestrial plants. These intercalations are continued towards 
the west. Von Ammon has described them in detail from the neighbour- 
hood of Partenkirchen in Bavaria^. At Seefeld in Tyrol petroleum has 
long been obtained from these fish-bearing partings, and the drops of 
' Saint Hubert's oil ' which are sometimes seen floating on the water of the 
springs come from these beds. On the Osterhom these partings are 
repeated several times in succession in the pale limestone ; a little higher 
and then between the beds of this limestone we perceive the first calcareous 
shales containing bivalves, such as in Swabia characterize the littoral 
development of the Bhaetic serie& 

These beds themselves have at first more or less the character of 
partings ; they increase in thickness, but the pale limestone continues to 
alternate with them in isolated bands; the whole series grows darker in 
colour, the clastic elements increase in importance; step by step as we 
ascend in the series, the mollusca of the deeper water preponderate more 
and more over those of the coast, until at length we find ourselves in the 
midst of the pelagic development of the Bhaetic series, yet once again 
the white limestone with Lithodendron makes its appearance in two beds, 
together 60 feet thick, and then finally the dark layers of the Bhaetic 

fairly deep into the cylindrical cavities of these corals. An analysis kindly made hy 

Dr. C. Natterer shows: 

Water (a) 0-04 (6)0-05 

Insoluble in hydrochloric acid 0*26 1-64 

Iron, traces of phosphoric acid 0«18 0-80 

CaCO, 55-76 55*10 

MgCOs 44.18 4276 

100-32 99-85 
The two beds, so different in appearance, are therefore both trae dolomites and differ 
only in the quantity of clay and iron they contain. ' 

^ L. Yon Ammon, Die Gastropoden des Hauptdolomites und des Plattenkalkes der 
Alpen ; Abh. zool.-min. Yer. Regensburg, 1878, XI, pp. 46-55. 



CH.VI] BONE BEDS IN THE PLATTENKALK 

Thna we start with the result that the Jiret Rhaetie tTUerccUaturt 
ths white IvmeatoTtea and dolomitea are not depoeUa of the de^ aea ; 
are charaeteriaed on the eontrary by a littoral fawna. 

Another example of the succession, quite as striking as th»t ol 
Osterhoro, has been described hy Zugmayr in lower Austria \ 

Close to the mill at Waldegg in the Piestingthal, where the thid 
of the Flattenkalk is eetimated as at least 1,000 meters, reddish 
appears as partings in the beds near the summit ; these fill up the tr 
depreesions in the upper surface of the beds, and sometimes unite tog 
to form Uiin continuous layers. They contain numerous scales, tee 
OyTole{H8, Sai^odon, Saurichthys, Acrodus, and other fishes, and rept 
the bone bed which accompanies the Rhaetie series in its littoral de\ 
meaat. They are repeated four or five times at least between the 
limestoae beds. Next follows hard limestone, again with Mega) 
Uien a parting nearly two feet thick, with Rhaetie moUusca of a lil 
character ; again limestone, and another reddish layer with a bone bed 
then, after some closing alternations, the representatives of the i 
Rhaetie sea, iacrectsing continually in depth. It would be superfiuo 
multiply examples. The presence of the marls and red shards, the hei 
together of the various kinds of limestone fragments, and perhapi 
the encrusting sinter, all combine to show that the surface of 
of the beds of the Flattenkalk was uncovered by the sea, expose 
a while to the air, and then again submerged. With the appea 
of the Rhaetie shells we observe an increasing proportion of de 
sediments. 

Id the snccebsicm of the Osterhom several facies may be distingu: 
which represent bo many different bathymetric stages in the Rhaetii 
The first of these is the Swabian facies: a bone bed, layers vrith M; 
and Taeniodon, with AvictUa contorta, bat without brachiopods. 
ancceeded by the Carpathian facies with Avicvla txmtoTta, Tefrebr 
gregarUt, Ostrea Saidingeri. Then comes tiie KSasen fades 
numerous brachiopods, such as Sjdrigeru osycolpos; and last of al 
Salzburg facies with Choristoceraa Marahi and Avicvla sp&Aota. 

The first three members of this series are found also at several loct 
in the north-eastern Alps. Schlonbach has described them at Kc 
they occur on t^e slopes of the Scesa plana in the Vorarlberg, as is s 
by collections which Esoher made in his time and presented to me, 
in other parts of Vorarlberg the Carpathian facies seems to dose the t 
Stoppani has traced these deposits with great care along the whole o 
limestone zone of the Italian Alps, and baa divided them into two groi 
a lower, consisting of shales and shelly limestones, the faciea of whi 

' K ZngnuL^ei', TJeber bonebedaitige Yorkomnmine im Dachiteinkalke dea Pii 
thalM ; Jahrb. k. k. geol. Reicha., 1875, XXV, pp. 79-88. 



266 MESOZOIC SEAS [PABTin 

pnrely Swabian ; and an upper, the beds of Azzarola, the fades of which 
is as distinctly Carpathian ^. 

The Swabian fades, maintaining a distinctly littoral character, extends 
far beyond the region of the Alps. At the most remote locality in Europe, 
Linksfield in Sutherland, it is only known in scattered blodu. From' 
Scotland it extends into the north-east of Ireland, and is continued thenoe 
across the counties of Nottingham, Warwick, Worcester, Qloucester, and 
Somerset, to Dorset. Some obscure littoral patches occur on the coast of 
Scania ; it is found over a large part of France, and the whole of central 
Germany. The Carpathian fades is far more restricted in its distribution. 
It scarcely extends beyond the region of the great folded mountains, but is 
known in the Jura, throughout the whole course of the Alps and the 
Carpathians, in the Apennines and in Corsica. It presents itself in typical 
development at Meillerie on the lake of Geneva (Favre), near the fall of 
Niinenen on the Stockhom (Brunner), and south of Hindelang between the 
nier and the Lech (Escher and Merian). The region of the Eossen fades 
is still more restricted. It is found in the north-east Alps, at several places 
in the Carpathians, and extends as far as Bukowina^ Sometimes the 
brachiopods of this zone lie together in dense masses in the light-red lime- 
stone of the Starhemberg beds, which occur only in the north-east Alps, 
as is the case also with the Salzburg fades, which has been less studied. 

Thus in the Bhaetic series of the Alps, the littoral beds occupy the 
lowest position, beneath all the other subdivisions ; at the same time they 
attain their greatest distribution over Europe. 

The upper beds of the Plattenkalk in the Osterhom were deposited 
during a series of oscillations with a preponderant podtive movement 
The preponderance, however, was so slight that the limestone was 
uncovered by the sea after its depodtion. It remained exposed to the air 
during the whole of the negative recurrence and the corresponding part of 
the next positive phase {^ef to -hkl, 11, p. 25). As the periods of sab- 
mergence became more and more prolonged, terrigenous sediment made its 
appearance, consisting, however, not of sand, but of day, and with it a 
littoral moUuscan fauna. With the persistence of the podtive movement 
the Swabian fades was replaced first by the Carpathian, and then by the 
Eossen facies, so rich in brachiopods. At the same time the whole of the 
central European sea advanced further and further beyond its previous 
bounda While, in the Alps, beds with the Eossen facies accumulated in 

^ U. SchlOnbach, Verb. k. k. geol. ReicliB., 1867, p. 211 ; A Stoppani, Easai ear les 
conditions g^n^rales dee couches k Avicula cantoria et snr la constitution g^ologique et 
paleontologique sp^ciale de ces mfimes couches en Lombardie, 4to, Milano, 1861, and 
supplement, 1868. Lepsius has described a Cyrena of the Swabian facies from the south 
Alps which points to fresh or brackish water; cf. his Das westliche Sild-Tyxol, 4to, 
Berlin, 1878, p. S60 ; the stratified succession, op. cit., p. 102. 

' Its occurrence near Taormina in Sicily has been recently disputed. 



ctt.vi] DEPTHS OP THE EHAETIC SEA 

deep water over the now completely buried Flattenkalk, eli 
the distant shores of England and Scotland, littoral beds n 
These contmnally advanced inland, characterized by the same 
the sune foasils as are to be found in the eastern Alps, whi 
either in the partings of the light-coloured Flattenkalk, o 
immediately above this stage. 

If we ascend the foot of the Ost«rhom to about 200 met 
first intercalation of the Swabian facies, we shall still find on 
econparatively deep-sea depoeitB, here represented by the 1 
brachiopods. The positive excess of the oscillations mnst 
have amonnted to more than 200 meters, otherwise none ba 
wtmld have been deposited on the site of the Osterhom. E 
+ <r--Neg.)>200 meters (II, p- 220), This positive movem 
itaelf in the littoral regions by a transgression which has I 
^bert and Martin over the whole of the south of the Ceni 
France, from the valley of the Rhone thxongh the departmes 
Qard, Loz^re, Corrfeze, and Dordogne, aa well as over the 
Nifevre, C6te-d'0r, and Sa6ne-et-Loire. This led H^rt to 
, the Central Plateau had undergone a gradoal subsidence i 
period '. In England, also, signs of this transgresi^ion are no 
it may have left its mark in still more distant localities. 

Thus, oa the lake of Geneva we find, according to Favre, t 
fades lying at a higher level than the Swabian ' ; In like manr 
the Bhaetic stage, where it is best known, as in Somersetshi 
borders of the Severn, as well as further north in Nottinghau 
of sandstone with one or more layers of bone bed, and it is n 
a little above this ihat shales, thin seams of limestone, and tl 
Swabian facies make their appearance ^ 

Our study of the numerous observations made on the sti 

* E. Hubert, Lee Hen ancieones et lenn riTa^rea dana 1« baarin de 
J. MartiD, Zone k Avieula contorla on £tage Bh^tien, Syo, Paris, 184 
^ 175. 

* A. ¥a.vm, M6moin tar Lei temitu Uaaqoe et keup^rien de la Savoie ; 
ethiftnat Oenfeve, 1859, XV, a, pp. 112 and 181. 

' T. Wright, HoDograph of the Lias AmmoniteB, Paiae. Soo., 18^ 
E. Wilson, The RhaeticB of Kottinghomehire, Quart Joorn. Qeol. Soo.j 
pp. 451-456. Ai early aa 1860 Hr. C. Hoore, of Ilininiter, vraa good eni 
til the Rhaetic epecimeiiB of hia collectioa, and Mr. T. DaTidaon, o 
Ki. T. Wright, of Cheltenham, have kindly furnished me with much ini 
■nbject. A journey to England completed my acquaintance with the 
eonntry. Mj object was to obtain a general idea of the Rhaetic sea. 
ledgo of the fextu served to enlarge the task ; my kind friends, Dan 
Wright, aa well as Etcher and Oppel, are long dace dead, and it is onl 
it Uat able to give here my final leenlta. Diadna Cdhntia from t 
lUiiaiell, and Ditdna Toum^imdi, Forbes, from Frome near Btiito 
Oppel alio sent me specimens of this ipeciea from Beit im Winkel. 






268 MESOZOIC SEAS [part ni 

sion and distribution of the Bhaetic series has now advanced- so far as to 
enable us to form some idea of the processes which accompanied the 
apparition of this new marine fauna, and to recognize and distingoish the 
signs of positive movement, on the one hand in the middle of the sea 
basins, and on the other in the littoral regions. At the same time we 
perceive that the dominant positive movement was not local, but extended 
as far as it is possible to pursue our investigations. Neither was it a sudden 
event, but oscillatory and slow. 

The presence of the red shards, which reveal the influence of the 
atmosphere, and the partings with fish and terrestrial plants have not^ 
however, hitherto been observed except in the Plattenkalk. We do not 
possess such evidence for all the beds, some hundreds in number, of which 
it is built up. It is possible that each bed corresponds to an emergence 
followed by a marine recurrence, but this is by no means certain. Indeed 
certain important facts are opposed to such a supposition ; as for instanoe 
the observations already made on the splitting of coal seams (II, p. 245)« 
The Carboniferous limestone beds are obviously formations of a similar 
kind to those of the Rhaetic epoch. The Fusulina of the upper Carboni- 
ferous limestone of the southern Alps have been drifted together into beds, 
in precisely the same way as the Qyroporella of the Bax-Alp. The paralic 
seams which occur between the beds of Fusulina limestone in the United 
States occupy the position of partings, and it is asserted of them that they 
unite together as they proceed landwards into larger, continuous, and 
probably autochthonous beds. Observations of this kind point far more 
to a periodic increase in the transport of sediment, than to extensive 
displacements of the strand-line. The tranquil deposition of the Oiannister 
series, the marine beds of no great thickness or extent in which the gaping 
bivalve shells have sometimes been found (II, p. 240), indicate on the 
contrary, as it seems to me, a transitory positive phase. 

Thus we are confronted by many an unsolved problem, and may almost 
envy our successors the discoveries which are reserved for them. 

The importance of a careful study of the partings has at all events been 
rendered clear. Walcott sees in the Galena limestone of Illinois and 
Wisconsin the representative of the XJtica slates, which were deposited in 
the east and north of the Silurian region of the United States; the transition 
is accomplished by the multiplication of the shaly partings in the lime- 
stone K The Devonian limestone Og^ of Prague is divided into numerous 
regular beds by clayey partings, and Elvaiia regards the superimposed 
Tentaculite shales Og^ as merely the continuation of the partings of Og^ \ 

^ C. D. Walcott, The Utica Slates and related FormationB ; Tians. Albany Instit., 1879, 
X, 88 pp., pi. 

' J. Klvi^a, Ueber die Silur-Schichten der beiden Moldau-Ufer, sCtdlich von Prag; 
Verh. k. k. geol. Reichi., 1888, pp. 87-43. F. Eatzer has subBequentljr examined these 



OSCILLATIONS IN THE LL4B 269 

becomee easy to trnderstond why Drasche, in view of the prevailing 
B OD coral reefs, was astoniBhed to find the coralliferoas limestone of 
which extends ap to considerable heights, divided into regular beds, 
b not to be wondered at that he Bhould even pat Qts question whether 
parent stratification, which also occurs in neta risiiig but a few feet 
the sea, might not result &om a periodic intermittence of the coral 

) Megalodus limestone and the Rhaetic setiee are not known outside 
ropean i^ea, except in the Himalaya and the outer chwns of the 
Kush, in particular at mount Sirban, near Abbotabad. In addition, an 
1 ^laetic fossil was found among the specimens brought home by 
from east Greenland. With the exception of the last-named locality 
aiae beds of the Bhoetic did not extend bey<md that re^on which 
I term, with Neumayr, the Central Mediterranean, or beyond that 
Ion of it towards the soath-west of Europe, which had already come 
istence as eu-ly as the Trias. 

nt-boEuing beds of Rhaetic age, on the other hand, are known in 
places ; e. g. in Siberia, Turkestan, Tongking, Australia, in the 
"ina and Karoo series, in the Argentine Republic, and in the eastern 
[ the United States. 

the'coast regi(m of the Pacific, again, no marine Rhaetic depoBits so for 
n aware have yet been observed. 

leed we have no knowledge at present of the occurrence anywhere 
oh marine fauna of this period. The Rhaetic beds with which we are 
nted contain a fauna of a somewhat restricted character. All those 
s groups of cephalopods which occur in the Trias and reappear in 
as should also be found in the Rhaetic, but strange to say they are 
id we know not where to search for them. When, at last, in the 
noBt port of the series a cephalopod does make its ^pearance, it is 
igular genus Choristiceras. 

Progreea of th£ podtive movement in the Juraagic period. We have 
led step by step hx)m the beds of the Flattenkalk through the 
ic oscillations, until we have reached a state of things in which open 
cupied the place of the eastern Alps, and the strand was gradually 
g outwards till it reached Scania and the north-east of Scotland. 
)W, we propose to follow the further oscillations of the Lias and the 
itages of the Jnraasic system. Neumayr, bringing to his assistance 
I recorded observations at his disposal, has attempted to follow over 

ud owing to the absence of Tentacnlitet in tbete intermediate bede baa not 

led this opinion, bnt the resnlte coincide aa regards the lack that the claj appears 

latod inteiraptions ; F. Eatser, Ueber wihieferige Einlagen in den Kalken der 

de'eehen Etoge Og„ Siteber. k. bOhm. Gee. Wiss., 1886, pp. 466-472. 

?on DiMche, Lnion, pp. 89 and 4S. The lame occurs on Cebd ; Abello, IbU de 

(.127. 



270 MESOZOIC SEAS pabtih 

the whole face of the earth the geographical changes which marked this 
period. The unexpected result of this laborious investigation was to show 
that to the Lias a very limited distribution appertains, while the approach- 
ing dose of the middle Jurassic is marked by a transgression of surpassing 
interest and extent \ 

It is dear, to begin with, that the oscillations persisted into the Liaa. 
Here the two regions in proximity to the ancient coast, u e. Scotland and 
Scania, again furnish important evidence. 

In Elngland, the further we advance towards the north, the more 
frequent become the intercalations of detrital and plant-bearing sediment 
in the Lias and other members of the Jurassic system, and they 
attain their greatest development in the isolated fragments which are 
preserved in Scotland, partly by faulting down, and partly by protection 
beneath basaltic sheets. Rightly to interpret these downthrown remnants, 
we must here, aa in the Rhine valley, first admit that these now isolated 
patches are the remains of a once continuous mimtle, which extended high 
over the whole of Scotland and its andent horsts. Judd had already 
adopted this view, as early as 1878, and was thus led by the same patii as 
Benecke in his study of Alsace, to a correct interpretation of the facts '• 

In the west of Scotland, as in the isle of Skye, for example, the lias is 
still complete, but the argillaceous partings of the limestone beds have 
become much thicker. To the north-east, in Sutherland, it is so no longer; 
detrital, plant-bearing beds have taken the place of the lower subdivisions; 
and only the upper part of the lower Lias and the base of the middle lias 
are represented by marine beds. It would seem, therefore, that it is these 
two latter horizons which attained the widest distribution in Europe. 

The case is the same in Scania; littoral beds of Rhaetic age are followed 
by a series containing shells, which is only slightly littoral in character, 
and these are the sole representatives of the basal part of the lias : the 
purely marine beds make their appearance on the same horizon as in 
Sutherland K 

^ M. Nenmajr, Bie geographische Yerbreitang der Joraformatioxi ; Denkachr. k. AkadL 
WisB. Wien, 1885, L, pp. 57-142, maps. 

* J. W. Jadd, The Secondaiy Rocks of Scotland, Qnart Journ. QeoL Soc., 1878, XXIX, 
pp. 97-197, and 1878, XXXIV, pp. 660-741, maps; for the west, also J. Btyoe, On the 
Jurassic Bocks of Skye and Baasay, op. cit, 1878, ^IX, pp. 817-851. 

' B. Lundgren, UndersOkningar Of^er Molluskfaanan i Syeriges ftldre MesoBoiika 
Bildningar; Lunds Univ. Arsskrift, XVII, 4to, Lund, 1881. The same hoiixons which 
have the widest extension in Scotland and Scania also seem to be encountered in the 
Alps as the material filling fissures in the Plattenkalk. Oxynoticertu oxynotu* is of 
frequent occurrence in north Scotland ; cf. Gteyer, Ueber die liasischen Cephalopoden 
des Hierlatz bei Hallstatt, Abh. k. k. geoL Beichs., XII, No. 4, 1886, p. 218 et aeq. ; by 
the same, Ueber die Lagerungsverh&ltnisse der Hierlatzschichten, Jahrb. k. k. geol. 
Beichs., XXXYI, 1886, pp. 216-294 ; F. W&hner, Zur heteropischen Differenzining det 
alpinen Lias, tom. cii, passim. The fissures described by Wright in the southeast of 
England are of different age, as are the Tertiary fissures m, m, Fig. 34, 1, p. 257. 



CH.TI] OSCILLATIONS IN THE JURASSIC 

We now enter the middle Jorassic In Franconia and Si 
marine sediments follow one another, bed after bed, in imdisturbi 
sion. This region was not subjected to negative recurrences. I 
of these are more evident in the north of France. Hubert comi 
trace tiieea out in detail as early as 1857, and he then recognizei 
Lias <m the Sarthe is incomplete, and that at various horizo 
Jurassic hard beds present UiemBelves with a water-worn, ai 
polishedt surface perforated by boring shelb, and he came to the i 
that the Juiasdc sea had enlarged its bounds by oscillatory movt 
to the epoch of the Eelloway stage \ la Normandy, which I 
1856 under the friendly guidance of MM. Deelongchamps, 
have been very exactly described by Eugtoe Deslongchamps. T\ 
inctHnplete, and its lower zones, speaking generally, appear to 1 
deposited in the north of the country only. Within the middle 
however, we may observe at various levels beds with a worn and i 
surface, known in the locality as 'chiens,' which indicate a t 
withdrawal of the sea. As early as 1864 Eug&ne Deslongchamp 
to conclude Uiat daring the Lias the sea in this re^on gradually 
its limits, and then, as the period approached its dose, witiidrew I 
distance; after this shallow water supervened, then another i 
(TrigoTtda navi«), followed by a fresh extension (Harpoceras Mur- 
Above the beds of Uie inferior Oolite a worn surface, drilled by Lit 
makes its appearance, and with this a new positive movement is 
which continued, advancing by slow degrees, up to the great 
sobmergence (Stephanoceraa ToacrocephalvM) ', 

The Jurassic deposits are continued across the Channel to Dc 
The marine series in England is at first fairly complete, bat tov 
ncKth there appear, as we have already seen, intercalated beds o 
material, and these continually increase in thickness as they appi 
mouth of some river in the north. In Lincolnshire the influeoc 
river has extended so far that one of the intercalations is to be se 
Dp channels of erosion in the underiying bed. In Yorkshire the 
of the detrital sediments has enormoosly iaoreased, and a consider 
of the marine middle Jurassic has been already replaced by them, 
describes eight small seams of coal, each with an underclay, into 

' Hubert aammed then that a retreat of the lea occurred after the EeU 
a negatiTe movement, but Uiia nev was probablj baaed on the appearance c 
lAich doretail (?) into one another, and owing to the general denudation of t 
form concentric lone* ; this ammgement was formerly regarded as an indical 
diminiehed exteniion of the na ; E. H^bett, Le« Hera anciennes et leura riva) 
lawin de Paria, 8to, Paris, 1857, The grey limeatoneB of Venice were obaei 
Sarthe ; Q. Boehm, ZeitMhr. dentscb. geol. Oe>.. 1887, XXXIX, pp. 205-211. 

* E. Ende»^Deiloiigchaffipi, Etudes snr les ^tagei jurasaiquei infSrieura < 
inni.i<i« ; H^. Soc. Linn. Norm,, 1865, XIT, pp. 1-296, in particolar p. 279 el 



5l5,-«^:r.ir: 



272 MESOZOIO SEAS [part m 

I 

is said to send roots ; these are interbedded in the plant-bearing series on 
the horizon of the Inferior Oolite. Finally, in the west of Scotland the 
whole series of the lower Oolites is resolved into a series of repeatedly 
alternating fluviatile and littoral beds, and it is not until the Oxf ordian 
stage is reached that a marine deposit makes its appearance. In Suther- 
land the marine development begins again with the Eelloway ^ 

Let us now turn our attention to the south. 

The deep borings made from time to time in London and its neighbour- 
hood have shown that in this locality the lias and Inferior Oolite are 
absent; the beds immediately superposed on the Palaeozoic formations 
begin with the Bath Oolite, which is followed by the Eelloway and the 
other marine stages of the upper Jurassic. It is precisely the same with 
the series superimposed on the Devonian reef at Marquise, near Boulogne, 
which also begins with the Bath Oolite (II, p. 92). 

Scania tells us nothing relevant ; since the beds above the middle LLaa 
have been removed by denudation, some loose fragments of the Eelloway 
have however been found. 

The scanty exposures of Jurassic which present themselves beneath the 
Quaternary deposits along the north coast of Germany reveal the presence 
of Lias ; and it has been reached by borings near Cammin. Still further 
east only higher beds are known, and in the boring at Purmallen near 
Memel the Eelloway beds were met with 95 meters beneath the surface, 
resting on red sandstone, which belongs probably to the Trias ^. From 
here the Eelloway extends towards Popilany, on the Windau in Lithuania. 

Thus the traces of a positive movement are manifest in the north of 
Scotland, beneath London, near Boulogne, and in the Baltic provinces; 
and we recognize a transgression beginning sometimes with tiie Oolite, 
sometimes with the Eelloway, and sometimes again with the Oxford stage, 
but in every case the Lias and Inferior Oolite are missing. 

We will return now to Franconia. 

The Trias deposits, as we trace them to the south, disappear on the 
western border of the Bohemian masa The lias is visible as far as 
Begensburg. As soon as we have passed the point where the Danube 
fault approaches the border of the ancient mass, that is, in the region of 
the marginal fracture which runs from Begensburg to Passau, the Jurassic 
series reappears, beginning with the lower stages of the Oolites. Beyond 
Passau every trace of the Jurassic has vanished, and it is not till we reach 
the far-distant locality of Olmiitz, north of Briinn, that we discover 

> J. W. Judd, The Geology of Rutland, Mem. Geol. Sunr. England and Wales, 8vo, 
1875, p. 33 et seq. ; A. G. Ramsay, The Physical Geology and Geography of Great BritaiB, 
5th ed., 8yo, 1878, p. 194. 

' G. Grewingk, Das Bohrloch von Purmallen bei Memel ; Sitsber. nai GeselL Doipat, 
105. Sitsung, 8?o, 1878. 



ca.vi] MIDDLE JURASSIC TRANSGRESSION ! 

lying apoD the Devonian a patch of Jurassic, the lowermost beds of wh 
according to Nenmayr and Uhlig, correspond to the upper Bathonian i 
the Kelloway *. 

Again for a great distance every trace of Jurassic is absent from 
border of the ancient mass, and it does not reappear until we reach 
neighbourhood of Cracow. Here and in the adjacent parts of Poll 
the stxatigraphic&l relations are very remarkable. The Trias is present ; 
Bhaetic series is represented in one locality by plant-bearing beds; 
lias is missing; sandstones and clays occur containing marine foesUs 
the lower stages of the Oolites. At Balin, west of Cracow, the faunas of 
Bath Oolite and Kelloway are richly represented, but it is the Eellov 
stage alone which is continued towards the east tc^ther witii the up 
series of the Jurassic. 

Next, let us enter Russia, and take for our guidance the works 
Nikitin '. 

The transgression begins in Poland with Uie lower Kelloway; t 
stage is found in the government of Kiev; in Ekaterinoalav, Kursk, i 
Orel, the Kelloway has not yet been subdivided ; the lower Kelloway t 
occnrs at EUatma in the government of Tambov, in Riazan, in the nor 
west of Simbirsk, and in parts of Nishni Novgorod ; it runs through 
east of Kostroma into the basins of the Vytchegda and the Petchora, t 
Qam reaches the Arctic Ocean. At the same time it extends through i 
east and south-east from Samara to Orenburg. 

C(mtemplating this long and comparatively narrow zone of lower Kellov 
we are led to ccmclnde that the sea must have entered the eroded basin 
a great river. The extension of the Jurassic sea did not ceaee with t 
stage, however ; the middle Kelloway extends still further, and enters i 
governments of Moscow, Tver, and Yaroalav. To the west of Kostroma I 
series begins with the upper Kelloway or the lower Oxfordion'. 
Kharkov the marine series, resting on plant-bearing beds, does not bef 
before the lower Oxfordian. 

We will now leave Europe, to examine the far-distant and isolal 
patches of Jurassic which are presented by the table-lands of Africa a 
India. 

' V. Uhlig, Die Jtuabildungen in der UniKebuiig von Brann; Beitr. pal. Gi 
0«st.-Ung., edited by K. v. Hojiiwvica and H. NeumaTr, iUi, 1882, 1, p. 180. Theie i 
the ' beds ofZeitlam ' of Uie neighbourhood of Ortenbnrg, at the uune time the equirale 
of the beds of Balin near Ci&cow. 

* S. Nildtin, Ueber die Beiiehnngen zwischen der ruraischen und der weatenropSiic) 
Joiftfomiation, Neues Johrb. Hin., 1886, 11, pp. 205-245 ; Neumajr, op. cit., 1867, 
pp. 70-88. 

' R Nikitin, Die Cephalopodenfanna der Junibildungen de* GonTemements Eostrot 
TerL niM.-k. min. Gea. St. Pelereburg, 1684, p. 74 ; and Allgemeine geologische Ea 
TOO RuMland, Blott 71, Hfra. Com. giol. Rutiiie, II, No. 1, 1885, p. 200. 



■i^-\-. 



274 MESOZOIC SEAS [pakt in 

At the south foot of the Great Hermon, near Medjel-eah-Shems, in 
Syria^ the Jurassic system is limited to a very restricted area. The lowest 
visible member corresponds, according to Noetling, with the lower Ox- 
f ordian \ 

The nearest locality at which the Jurassic is next known to crop out 
from beneath the great mantle of Cretaceous limestone is Antalo, in the 
north-east of Abyssinia (I, p. 868). Aubry, who has somewhat recently 
travelled through the region in which the Blue Nile takes its rise, has 
furnished us with a detailed description of the geology. The sheet of 
Cretaceous limestone which covers the north-east of the Sahara does not 
extend into Abyssinia. It doubtless comes to an end near Khartoum, and 
the Archaean foundation is exposed throughout the Soudan, and according 
to Schweinf urth as far as the land of Niam-Niam. On the coast region it 
is visible as far as Massowah. This Archaean region is crowned by the 
high plateau of Abyssinia, consisting like the Sahara of flat-lying beds, bat 
with a different stratified succession. The lowest member of the series is 
the thick mTidsUme of Adigrat, which extends in an arc from the foot of 
the table-land, probably starting from Adigrat and entering the interior as 
far as the sources of the Blue Nile. It is white or sometimes bluish, and is 
overlain by a yellowish crystalline limestone, which contains little bivalve 
shells and intercalated gypsum and dolomite. This is succeeded by tiie 
AntaXo limestone, of Jurassic age, with marine fossils; Blandford dis- 
covered it at Antalo ; it is visible again on the slopes of the valley of the 
upper course of the Blue Nile, and probably also in the valley of the 
Ouibi^ in Eaffa. Above this rest beds of sandstone and gypsum, not so 
thick as the preceding, and the simimit of the table-land, 2500 to 2800 
meters above the sea, is formed of vast outflows of lava \ 

It is to this mantle of volcanic rocks that we are indebted in all 
probability for the preservation of the Jurassic succession in Abyssinia, 
represented in patches which are probably the remnants of once very ex- 
tensive beds, and still even at the present day prolonged fairly far to the 
south. The marine series, according to the comparative studies of Douvill^ 
again begins with the Bathonian stage, and probably extends into the 
upper Jurassic as far as the horizon of the Eammeridge. 

To the southern extension of the Abyssinian series we may perhaps 
refer the exposures of upper Jurassic made known by Fraas and Beyrich 
at Mombassa on the coast of Suaheli (I, p. 400). 

Another patch of flat-bedded marine Jurassic occurs at Cutch in India, 
there also partly protected by a more recent mantle of lava. It is oon- 

^ Diener, Libanon, p. 25 ; F. Noetling, Der Jura am Hermon, 4to, Stuttgart, 1887* 
* Aubry, Observations g^ologiques sur les pays Danakils, Somalis, le rojaume da Choa 

et les Pays Gallas, Bull. Soc. geol. de Fr., 1886, 3^ s^r., XIY, pp. 201-222, geol. map ; 

H. Douvill^, Ezamen des fossiles rapport^s du Choa par M. Aubry, tom. cit, pp. 223-241. 



CH.VI] EXTENSION OP UPPER JUEASSIC SEA 275 

tinued in a series of outliers far to the north-north-east in the plain of 
Rajputana (I, p. 414). Waagen's investigations show that the lowest 
member of this marine series, the Putchimi group, also corresponds to the 
Bath Oolite; Douviiy has pointed out the resemblance of the Antalo lime- 
atone of Abyssinia to this group. Above it comes the Eelloway stage with 
Siephanoceras macrocephaluTn, and then a further series of marine stages 
presenting a most remarkable correspondence with the upper Jurassic of 
Europe. They are probably continued into the Salt range ^. 

It is unnecessary to repeat what has already been said above regarding 
the less exactly known beds, probably of middle Jurassic age, which occur 
in Madagascar and on the east coast of India; I will only recall the fact 
that in West Australia the lowest marine deposits so far known also belong 
to the middle Jurassic (II, p. 189). 



Golgi^ 



Abal or the Blue Nile. 




Fio. 28. The vaUey <f the Blue Nile (after Aubry). 

I. Yoleanic roeks, 800 meters thick ; s. YeUowiBh crystalline limestone with Trigonia, 100 
meters ; 3. Grey marly limestone with Modioia oigMra, Terebratula, Geromyai 400 meters ; 
4. TeUoiilBh crystalline limestones, with layers of dolomite, gypsum, and small bivalTe shells, 
800 meters; 5. White and blue sandstone, sometimes micaceous, with intercalations of 
g:reein and yariegated clay, 600 meters. 

The results of our inquiry may be summarized as follows : — 
At the beginning of the Bhaetic period the sea, in the region we have 
had under consideration, occupied a narrow area confined to a part of the 
Alps and the great ranges of Asia. Its level rose in Europe by a series of 
oscillations ; its. waters extended finally over a great part of central Europe. 
The oscillations persisted throughout the Lias ; at the opening days of the 
Idas the domain of the sea was diminished, it was then enlarged and again 
reduced. It gained nothing in the lower stages of the middle Jurassic ; 
but, with the advent of the Bathonian, which no doubt is not over sharply 
separated by its fauna from the Eelloway stage, the strand-line rose far 
and wide. The Bathonian lies on the downthrown Armorican arc beneath 
the soil of London, and on the Devonian at Boulogne ; with it commences 
the great transgression over Abyssinia and the north-west of India. The 
boundary of the sea enlarged still further during the deposition of the 

* W. Waagen, Jurassic Fauna of Eutch : I. The Cephalopoda ; Mem. GeoL Sunr. India, 
ser. IX, 1, 1873, p. 225. 

T2 






p. 






I - 



M 



► / 



I 






\ 1 
I 



I • 






276 MESOZOIC SEAS [PABim 

Eelloway ; in the extreme north of Scotland this lies on the flaviatile oosl- 
bearing beds of Sutherland, it extends over the lower Jurassic stages of 
Pomerania, proceeding towards Memel and as far as Lithuania ; and while the 
Idas of Franconia already comes to an end at Begensburg and the other stages 
of the middle Jurassic disappear in their turn, the sea of the lower Kellow&y 
stage extended from Poland past Kiev, forming a long belt on the western 
side of the Ural and so reached the Arctic Ocean; at the same time, passmg 
Orenburg, it encroached on the eastern side of that chain. At this epoch 
the Jurassic sea extended from the Petchora on the one hand to Sutherland 
on the other, to Abyssinia and Cutch, and much further still to the south 
and south-east. Withal the transgressive beds have preserved a hori- 
zontality as undisturbed on the banks of the Petchora as on the Blue Nile ; 
and Stqphanoceras ma>crocephalum maintains its horizon in the Eelloway 
from Brora in Sutherland to Cutch in India. 

This extraordinary extension did not, however, mark the culminaiaoii 
of the positive movement ; the coasts became increasingly remote, and this 
renders the determination of the new transgressions more and more diffi- 
cult. The succeeding stages occupy, as far as can be seen, not only the 
whole region covered by the Bathonian and Eelloway, but in Europe they 
even extend beyond it. 

Near Olmlitz the Eelloway stage and lower Oxfordian reached the 
summits of the Devonian zone of the Sudetes ; the upper Oxfordian and 
the lower Eimmeridgian ascended over the summit of the Bohemian masB 
and advanced across it into Saxony. Some slight remains of these trans- 
gressive deposits are now to be seen pinched in by the reversed fault whidi 
borders the southern edge of the Riesengebirge and the Isergebirge (I, 
p. 212). 

In like manner, during the Eimmeridge epoch the sea covered that 
southern part of the Bussian Platform which is cut through by the upper 
Dniestr (I, p. 181) in eastern Galicia, and extended over the Devonian red 
sandstone at the same time ; it also reached the Dobrudja and deposited the 
horizontal limestone beds of the Eimmeridge on the upturned green 
schists of the mountain fragments of Matchin, Hirsova, and Chemavoda, 
extending as tax sia the Black sea; the limestone beds in all probability 
proceed still further beneath the Bulgarian plain (I, p. 476). 

The investigations of Russian geologists lend support to the conjecture 
that in this case also the transgression of the lower Eelloway does net 
mark the highest level attained by the strand. We have already men- 
tioned the later transgressions which seem to have occurred during the 
upper Eelloway and the Oxfordian. Whether the Eimmeridge stage 
covered great surfaces in Russia as it did in Oalicia and on the lower 
Danube, extending beyond the limits of the Oxfordian, is however not 
known. This stage, as we shall see directly, marks the last recognized sub- 



OH.TI] RECESSION AT CLOSE OP JUR. 

margenee known to have occnrred before a great chao) 
in, and it may therefore have been expoeed to denndati 
decree than the other stages. Pavlow, who faaa rendei 
aerriee hy hia atadiea of the Eimmeridge in Russia, desc 
at several localities in Simbiisk, as well aa in the neigh 
boi^ ; it also oocon according to Qoorow on the shores 
aeeording to Levisson-Leesing at Nishm Novgorod. 1 
preciaely the same characters as in the contemporaneous 
Etixt^ ; SkBogyra virgvla, the little oyster so abondan 
have led to the creation of an independent sab-groap, t 
well known also in the Eimmeridge of Spain, Eiiglan< 
north of France and Hanover, as well as in the platy 
now likewise appears in Poland and in the south-east 
pany with many other characteristic speciea of the Eii 
western Europe *. 

With the close of the Kimmeiidge a complete chan^ 
in and affected the whole of Europe. Everywhere th 
Bnasia a new marine fauna, alien to the west of Europ 
aooe, advancing from the north ; it marks the Volga « 
case is similar. Above the Katrol sandstone, which < 
BoTopean Elimmeridge, plant-bearing beds occur; thee 
Ettr(^>eaD fauna present themselves, proceeding out of th 
hnna of the Uitenhage serUa of southern Africa. 

We will now briefly review, so far as the evidence 
very incomplete, permits, the emergence of a part of c 
then the fresh tranagreesions, first coming from the norl 
from the south. 

4. Negative phaee avd the begvnning of the Cretm 
extttit and thickness of the fresh-water deposits whicl 
at the limit of the Jurassic and Cretaceous systems ar 
have given rise to frequent discussions of a general 
R. Oodwin-Austen, in giving a general account of these 
very suggestive remarks, which strongly recall those idee 
of sediments in cycles to which we have made frequi 
believed that we coold recognize in the northern hemit 
of the same physical ctmditions after extremely long 
The marine series is again and a^;ain interrupted, ov 
continental deposits, like the Old Bed strndstone for 
meaearee of the Carboniferous system, and the freah-wf 
Weald. 

' A. PktIow, Note tur lluatoire de la fiuine Eimmeridgienne i 
HoKon, 8vo, 1886, 16 pp. ; by the Mme, Lea AmmoniteB de li 
aeanthicam de Veatde la RoMie, H^m. Com. g6ol. Ruuie, II, No. ] 



278 MESOZOIC SEAS [PArani 

In support of his views Godwin- Austen appealed to several of the 
Wealden areas. A great lake covered the south-east of England, extending 
over Kent, Sussex, and Hants ; it was continued across the Channel into 
the Boulonnais, and has left its traces in Oxfordshire and the isle of Wight 
Oodwin-Austen gave as the minimum size of this lake a length of 190 
kilometers taken from Portland to beyond Aylesbury, and a breadth of 
320 kilometers from Portland to the Boulonnais. Taking into account some 
small patches of Wealden in the anticlinal of the Pays de Bray, we obtain 
for the line from Wiltshire to beyond Beauvais a length of 400 kilometers. 
Other traces of the deposit have also been found in the departments of 
the Aube and the Jura. In the Charentes another lake existed about 
as large as lake Ladoga. A third in north Germany, about 190 kilometera 
in length, reached from Ibbenbiihren through Osnabriick to Hanover 
and beyond. Finally, certain isolated patches of plant-bearing beds lying 
on the Belgian coal-field are also assigned to the Wealden series^. 

Subsequent observations have brought to light many new facts, and 
with their aid I will now attempt to trace in broad outline the course 
of events in western Europe. 

The sea reigned over the whole region of the Alps and the Carpathians 
during this period «. Deposits representing the Weald are known in Spam 
and Portugal, and we must not omit to take them into account. But we 
will consider first the regions best known ; these extend from the south- 
east of England to the Jura mountains, and from the west coast of France 
to the Harz. We shall depend for our most important data on Judd, 
Meyer, and Topley in England ; H. Credner and Struckmann in Germany ; 
Pellat, Loriol, Cotteau, Jaocard, and Maillard in the east of France and the 
Jura ; and on Coquand in the west of France ^. 

* R. God win- Austen, Address to the Geological Section ; Rep. Brit Assoc. Brighton, 
1872, pp. 90-96. 

^ The only exception is perhaps mont Salvens, a chain of foothills of the Freiburg 
Alps, where blocks in the lower Neocomian and denudation of the upper JuraMic 
limestone beds may be quoted as indications of an interruption; V. Gillian, 
Aper^u g^ologique sur les Alpes de Fribourg, Beitr. geol. Karte Schweiz, XII, 1873, 
p. 108 et seq. 

» J. W. Judd, On the Punfield Formation, Quart Joum. GeoL Soc, 1871, XXVII, 
pp. 207-227 ; G. J. A. Meyer, On the Wealden as a fluvio-lacustrine formation, op. cii, 
1872, XXVIII, pp. 243-255, and op. cii, XXIX, 1878, pp. 70-76 ; W. Topley, The 
Geology of the Weald, Mem. Geol. Surv., 1875, 8vo ; H. Credner, Ueber die Gliedening 
der oberen Juraformation und der Wealdenbildung in N.-W. Deutschland, 8vo, Frag, 
1868, et passim ; C. Struckmann, Die Wealdenbildungen der Umgegend von Hannover, 
Hannover, 1880 ; by the same, Ueber den Parallelismus der hannover*8chen and der 
englischen oberen Jurabildungen, Neues Jahrb. Min., 1881, pp. 77-102 ; by the same, 
Die Portland-Bildungen der Umgegend von Hannover, Zeitschr. deutsch. geoL Gee., 1887, 
XXXIX, pp. 82-67 et passim ; P. de Loriol et E. Pellat, Monographie pal^ontologiqne et 
g^logique de I'^age portlandien des environs de Boulogptie-sur-Mer, M^m. Soe. phys. 
hist. nat. Geneve, 1866, XIX, in particular p. 186 et seq. ; P. de Loriol et G. Gotteao, 






•'.. ,„■ 



CH.VI] GYPSUM OF UPPER JUEASSIC 279 

(a) The Kimmeridge stage extends from Portugal through the whole 
of the region considered here as far as Orenburg. 

(6) In western Europe the Eammeridge is succeeded by another marine 
member of the Jurassic, the Portland stage. This always lies conformably 
oa the Kimmeridge, but in Hanover, where its fauna is very varied^ 
74 per cent, of the lower Portland species are known in the preceding 
Kimmeridge beds. The limit between the stages is by no means well 
defined ; some geologists do not admit its existence ; the course of events 
was therefore fairly undisturbed. In the upper beds of the Portland the 
Gomiexion with the Kimmeridge is less marked, and in England, as in 
Oemany and the Jura, brackish-water species, or such as are less sensitive 
to a change in the composition of the water, to a certain extent pre- 
dominate. In the Charente, as in England, Hanover, and the east of 
France and throughout the Jura, the shells of Corbula inflexa and Cyreria 
rugosa cover in millions what was once the flat bottom of a shallow sea, 
apparently somewhat less salt than the Ocean« 

But on the upper Dniestr we again discover a whole series of Portland 
fossils, including abundant examples of Corbula infleaxi; and the facts 
elicited by Alth in his study of this fauna forbid us to assume that 
a barrier separating eastern and western Europe was in existence at 
this tinie^. 

We have now reached the upper limit of the Portland represented by 
the beds which are known in Hanover as the ' Einbeckhauser Plattenkalk.' 

At this point the great change set in. 

(c) The sea, which extended beyond the Volga at the time of the 
Kimmeridge, and at least as far as the Dniestr during the Portland, was 
now so closely restricted that it did not pass beyond the region of the Alps 
and Carpathians, and indeed did not retain its hold on the Jura. In the 
territory it had abandoned lagoons remained in which clay and gypsum, 
with here and there rock-salt, were deposited. 

In EIngland these gypsiferous beds, which are always very poor in 
organic remains, are never exposed at the surface; but when in 1874 

Monographie pal6ontologique et g^logique de T^tage portlandien da d^partement de 
I'Yonne, Bull. See. sci. hist. nat. Yonne, 1868, 2* s^r., I, in particalar p. 241 et seq. ; 
E. Pellat, £meiBion da sad et de Test da basdn pariden & la fin de la periode jurassiqae, 
etc., Bull. Soc. g^l. deFr., 1875-1876, 8« b^., IV, pp. 364-369 ; by the same, Le Terrain 
joianiqae mojen et sup^riear du Bas-BoulonnaiB, op. cit., 1879-1880, 3* s^r., VIII, 
pp. 647-699 ; Loriol et A. Jaccard, £tade g^blogique et pal^ontologiqae de la formation 
d*eaa donee infracr^tac^ du Jura et en particulier de ViUers-le-Lac, M^m. Soc. phys. 
hist. nat. Geneve, 1865, XVIII ; G. Maillaid, Invert^bres da Purbeckien du Jura, M6m. 
Soc pal^nt Soiflse, 1884, XI, 156 pp., map, supplement in 1885, XII ; H. Coquand, 
Description physique, g^ologique, etc., du d^partement de la Charente, 8vo, Be8an9on, 
1, 1858. 

^ A. von Alth, Die Versteinerungen des Nizniower Ealksteins ; Beitr. palftont. Geol. 
Oesterr.'Ung., edited by Mojsisovics and Neumayr, 1882, 1, p. 830. 



/ 



* 



280 MESOZOIC SEAS [PAimn 

a boring was made near the middle of the Weald, they were encountered 
with a thickness of over 100 feet. They lie at the bottom of a basm, 
probably as a lenticular mass, completely buried out of sight beneatii 
more recent sediments \ 

In north Germany they are known as the Munder marls, and they are 
frequently exposed at the surface as a consequence of the folding which 
has affected this region. The salt-beds underneath the city of Hanover, 
which were passed through by a boring, belong to this serie& The maris 
attain a thickness of over 800 meters. The resemblance they bear to the 
Eeuper has been pointed out by Struckmann. 

A map of the Purbeck deposits in the Jura constructed by Maillard 
shows clearly that the course of events was the same there as in EngUmd. 
Oypsiferous marls, but only 3 or 4 meters thick, were deposited above 
the Portland over a part of the region now occupied by the Jura mountains, 
chiefly between the Doubs and the lake of Neuch&tel, and somewhat 
further to the north-east and south-west; but the succeeding deposits 
have a much vnder extension, particularly towards the south-west, and if 
the region had not been folded, the gypsum would form a lens buried out 
of sight as in England. 

Far to the west also, on the Charente, the gypsum has been laid 
down over the Portland. In this case, we have before us only part of 
a basin ; the rest is concealed, on the one hand by the Atlantic Ocean, 
on the other by the transgression of the middle Cretaceous. It corre- 
sponds to the southern border of the Armorican horst, which is followed 
by the several members of the Mesozoic series in such a manner that 
successively younger members make their appearance as we proceed to the 
south or south-west. Above the Portland lies first a bed of cavernous 
limestone, about 1*6 meters thick, then 35 to 40 meters of gypsum accom- 
panied by clay, with fish scales and fragments of wood. This bed of 
gypsum crops out in a zone striking to the north-north-west, from 
Ch&teauneuf west of Angouleme for a distance of 40 kilometers, and it 
may be traced in isolated exposures through Bochefort to the point de 
Chassiron, the northernmost promontory of the island of Ol^ron ; so that 
its total length is more than 100 kilometers. The Gironde and the Atlantic 
conceal the greater part of this basin, and its outline is unknown. 

The remarkable negative TnovemerU, which irUroduced such a ccymjMe 
change of conditions, also brought about the separation of eastern fnm 

' It is trae that by yery close observation ve may recognize the border of the synclinal 
of gypsum, sometimes indicated by corrosion of the sor&ce of the Portland, sometimes ai 
a thin layer of clay, or otherwise ; see Blake, Quart. Joum. QeoL Soc, 1880, XXXVI, 
p. 221 ; Andrews, op. cit., 1881, XXXVII, p. 249, also Girardot, Bull. Soc. g^ol. deFr., 
1884-1885, 3* s^r., p. 755. Perhaps the so-called middle Portlandian of Boulogne is 
to be referred here. 



WJ*f #^ ■ S I * |p.i|]f}Jip.l ^# ^Ji 



•1 ■ 



CH.VI] IMPOVEEISHED JURASSIC FAUNA 281 

weeiem Europe, and for a considerable interval each of these regions was 
the scene of independent events. This episode corresponds with the lower 
part of the Furbech The division of Europe marks its advent. 

{d) The basins in which the gypsum was deposited began to be refilled. 

In the Jura the sediments of this new stage, which consist of limestone 
and marl, extend from Biel to the south-west through the whole mountain 
range and further still into the department of the Is^e. They are some- 
times separated from the underlying beds by a cavernous dolomite, and 
attain a thickness of only 4 or 5 meters. At their base is a layer con- 
taining a mixed fauna of fluviatile and brackish- water shells, then a fresh- 
water bed, and above that again brackish-water layers. The upper 
brackish- water bed is, however, of but trifling extent, and appears, at least 
according to existing observations, scarcely to pass beyond the region of the 
gypaiferous clay, which lies deep below it. The fresh-water limestone-beds 
reappear outside the Jura near Baume on the Doubs, above Besanfon, and 
near Gray on the Sadne, so that a connexion appears to have existed with 
the eastern side of the Paris basin. But in the department of the Tonne 
the beds above the Portland are already absent, as well indeed as every 
trace of fresh- water deposits ; and the marine lower Cretaceous rests directly 
on the marine Jurassic. 

In north Germany, an imposing series of limestone beds, with marine, 
brackish, and fresh-water shells, was deposited during this period above 
the marls of Munde : they are known, owing to the abundant occurrence of 
Serpula coacervatay as the 'Serpulite.' The Serpulite also occurs near 
Boulogne and corresponds, in England, with that part of the Purbeck 
which lies above the gypsum. Oscillations were so frequent that Bristow 
distinguishes in this zone as developed in the isle of Purbeck, the remains 
of four ancient forests, eleven fresh-water beds, four brackish, and three 
marine beds, succeeding one another in repeated alternations. 

On the Charente a limestone bed with fresh- water moUusca also occurs, 
the * couche de deux pieds.' 

The marine or brackish- water moUusca of these beds fully maintain the 
character of the lower beds ; many species are common to both ; we have 
here an vmpoverished Jurassic fauna. 

The later transgressions show clearly that at this time all the western 
part of the Paris basin beyond the Pays de Bray was dry land, while in 
the east, from the Alps to England, the marine and fresh- water faunas 
variously alternate with each other. 

(e) While the course of events had hitherto been uniform over vast 
areas, now local differences began to make their appearance. 

The strand rose still further ; the regions lying nearest the Alpine sea 
were submerged, and the new marine fauna of the Yalengian or lower 
Neocomian advanced from the south over the region of the Jura. The sea 



282 MESOZOIC SEAS [pabtih 

penetrated into France, but towards the west its deposits do not extend far 
past Vailly, north-west of Saneerre in the Cher. In the north of FrwMe, 
however, and vn the whole of England the mariTie deposits of the lower 
Ifeocomian are not known to occur \ 

It is true the sea did not at once enter into permanent possession. Here 
and there occasional intercalations of fresh-water limestones may be met 
with, but these at length cease as the sea acquired undisputed dominion. 
Somewhat further north, in the Haute-Mame, a fresh-water bed is seen 
again at a much higher level, the Urgonian stage of the Cretaceous. 

Still further north, in Hanover and in England, fresh water persisted 
longer than in the neighbourhood of the Alps ; indeed it was at this period 
that the typical deposits of the Weald first began to be formed. 

In north Germany the Serpulite is covered by a deposit of sandstone 
and shale, 200 meters thick, containing the remains of reptiles, fresh-water 
moUusca, such as Unio and C3nrena, and beds of coal, many times repeated 
in succession, as in the ancient Coal-measures. Tracks of animals on the 
surface of the sandstone beds show how the surface of the fresh- water lake 
gradually rose to a higher level, till it became stationary. 

Similar sandstones occur in small patches near Boulogne ; in England 
they lie above the Purbeck, as in Hanover, and contain numerous remains 
of the great Iguanodon and fragments of terrestrial plants. The series is 
known as the HaMings sands, and, as in Germany, the footprints of anima l s 
are found in it at various levels. 

(/) In Hanover, as in England, there now follows the upper Weald day. 
Again the influence of the sea began to make itself felt 

In north Germany this member attains a thickness of about 60 to 70 
meters: by the side of numerous fluviatile mollusca, the brackish- water 
forms of the Serpulite reappear, and among them Corbula inflexa. This 
represents the still further impoverished Jurassic fauna, which during the 
period of the Hastings sands had continued to exist somewhere in lagoona 
In England we meet with brackish- water intercalations in the fresh-water 
clay, and Ostrea distorta once more gives rise to a bed which occurs dose 
to the summit of the Wealden. Thus under the most unfavourable con- 
ditions the last survivors of the Jurassic period have succeeded in main- 
taining their existence, a true relict fauna. 

In England the thickness of the deposits increases rapidly towards the 
west, and reaches 2,000 feet in the beds above the Purbeck : possibly the 
clastic sediment came from that direction. In Belgium there may be seen 

^ Hubert has done admirable work in establiihing these facts; an accoont of the 
conditions in France is given by Lapparent, Traits de g^ologie, 2* ^, p. 1084 at seq. 
Price (Proc. Geol. Assoc., 1875, IV, pp. 269-278) and Gardner (Geol. Mag., 1877, 2nd ser., 
TV, p. 377 et passim) conclade from the nature of the deposits that the Ganlt sea of 
England also became deeper. 



CH. VI] DISAPPEARANCE OP WEALDEN LA 

here and there, resting on the Coal-measures, beneath the h 
middle Cretaeeotis tran^ression, patches ot sand or clay 
remains of plants. These are the remnants of depoaite whi 
in hollows of the ancient surface; in 1878 the mines of i 
near Bemissart, not far from the French frontier, revea 
of S22 meters below the ground, a mass of clay faulted dowi 
measures ; this contained several skeletons of Iguanodon, 
fishes and plants of the same species as characterize the Enf 

Brown clay containing plant remains occurs also on t) 
Charente. 

(g) The sea advanced again still further to the north, 
of great lakes gradually came to an end. It is no longer tb 
of the Valengian or lower Neocomian which accompanies tJ 
as in the Jura ; but that of the middle Neocomian, the Hils 
as we meet with it in Germany, At the same time the n 
was submerged, and the south of England was reached by 
tiist marine intercalation of the Cretaceous lies very nearly 
But in England fresh water once more reappeared, and it i 
following stage of the lower Cretaceous that the region of t 
was at length permanently occupied by the sea. This sta 
the Pnnfield beds, which correspond to the Urgonian of Frei 

The positive movement proceeded in a series of oscilli 
carried the lower Neocomian from the Alps across the reg 
to France. In the middle Neocomian, ie. at the time of 
a further transgression brought its waters over the north 
north of Germany, the north German Weald, and a large p 
In the following or the Urgonian stage, the series of fresh 
was brought to an end in England also. While the se 
Cretaceous had already spread itself over the Purbeck ar 
some brackish-water mollusca, the last remains of the Jrni 
persisted in the Wealden area of England and Germany, dei 
wards march of the sea °. Indeed, these Jurassic mollusca, i 
by repeated oscillations to brackish-water conditions, we 
fitted to maintain their existence in the lakes and estuarii 
than the new arrivals freshly introduced by the waters of 
sea. 

Having obtained these results, let us now east a g 
peninsula beyond the ^renees. 
' Dapont, BoU. AcaA Boy. Belg., 1878, 2" aft:., XLVI, p. 887. 

* In Germanj also some fresh-water mollnBca occur in higher 
G. Boehm, Beitr&ge mr geoguoBtischen Eenntniss der HUtmnlde, Zelt 
Gei., InaagTual-DuaertBtion, Berlia, 1877, p. 10. 

* Thia important conclusion was also drawn by Bejrich, Zeitschr. ( 
ISSOi and Nenmayr, Terh. k. k. geol Reichi., 1880, p. 290. 



284 MESOZOIC SEAS [PAmm 

We have found hitherto that we could most clearly represent the course 
of events by tracing them from the Alps through the Jura and on to the 
north-west as far as the English Weald. We will pursue a similar course 
in Spain, and follow the succession of the beds in a north-west directioii 
from the Balearic isles to the Atlantic coast. 

In the Balearic isles, which we regard as the probable continuation of 
the Betic Cordillera, the stratified series, as in the cordillera itself, and as in 
the Alps, is chiefly of a pelagic character. Purely marine deposits repre- 
sent the upper Jurassic, which is overlaid by the lower Neooomian. On 
the continent, we again meet with the lower Neocomian near Valencia, but 
further inland it is not yet known to occur ; it is also absent in those parts 
of the Paris basin which are most remote from the Alps. On the other 
hand there appears on the coast, in the little mountain range of Atalayas 
de Alcala, near Castellon, a lithographic limestone with Exogyra virgvltti 
and thus on the horizon of the Eimmeridge ; it resembles the lithographic 
limestones which occur at the summit of the Jurassic in the south-east of 
France, and from Ulm to Solenhofen in Bavaria, but these probably 
represent the Portland as well as the Purbeck, and GUmbel has shown that 
in south Germany they bear the signs of numerous minor oscillations, such 
as under different conditions we were able to recognize in the English 
Purbeck, The next sediments which are known above the lithographic 
limestone of Atalayas belong not to the lower Neocomian, but to the 
Urgonian stage \ 

Further inland in Teruel, coal beds crop out, alternating near Utrillas, 
with marine sediments ; their fauna precisely coincides with that of the 
Punfield beds, which are intercalated with the highest beds of the English 
Weald, and mark its close '. 

The fresh- water deposits of the Weald occur still further to the north- 
west in parts of LogrofLo and Burgos, extending as far as the valley of the 
Saja» in Santander, and attain the Atlantic coast. Palacios and Sanchez 
have described them in the southern part of this region ; according to 
Calderon's account they attain a thickness of over 1,000 meters, and the 
largest of the areas they occupy covers 1,200 square kilometers ; they may 
be subdivided into Hastings sand and Weald clay, as in England ^ 

^ H. Hermite, J^tudes g^ologiques gur les lies Bal^res, 8vo, Paris, 1879, p. 138; 
Coquand, Sur I'ezistence des stages coraUiens, etc., dans la province de Castellon de 1a 
Plana, Bull. Soc. geoL de Fr., 1867, 2« s^r., XXIV, pp. 462-471 ; C. W. Gfimbel, Die 
geognostischen Verh&ltnisse des Ulmer Gementmergels, Sitzber. k. bajr. Akad., 1871, 1, 
pp. 88-72, 

> £. de Yemenil et 6. de Lori^re, Description des fossiles dn N^oconden sup^enr 
d'XJtrillas, 4to, Paris, 1868 ; the correspondence is emphasized by Jadd, Quart Joain. 
Geol. Soc, 1871, XXVII, p. 224. 

' P. Palacios J Raf. Sanchez, La formacion Wealdense en las provincias de Sorio y 
Logro&o, Bol. Com. mapa geol. Espana, 1885, XII, pp. 109-140 ; S. Calderon, Note sor 



CH.VI] WEALD LAKES OF SPAIN AND POKTUG. 

The manne seriea of the Cretaceous begins everywhere in t 
Spain with the Urgonian, which, as Carez has shown, extends over 
north Arragon, Navaira, and Biscay K 

The correspondence with central Eozope is remarkable ; tb 
isles answer to the Alps, Santonder to the soath of England. 

Let as next torn to Portugal. 

Many years ago Sharpe described an ' infra-Cretaceous ' serie 
water ongin occurring in Portugal. Delgado recognized its com 
with the Weald, and it was sabaequently described by Rib< 
shown on the geological map of Portugal as extending from 
Cabo de Ejspichel, and in many larger and smaller areas north of 
to Cabo Mondejo. Saemann correlated the upper parts of the ] 
Jnraesic with the Portland of Boulogne '. 

In the Serra de Cintra, acconUng to Choflat, the Eimmeric 
moonted by a thick series of limestone beds, which contain sa 
species of the preceding fauna, and some fresh arrivals, includis 
llien follows Uie Valengion, next the Hauterivian, i. e. the ma 
aod middle Neoc<nniau. But only 20 kilometers to the east, m 
weet-ntnih-west of lisbon, the Valengion is replaced in this s 
sandstone containing terrestrial plants and corresponding to the ' 

The aeries of events is tima essentially the same from the H 
west coast of PortugaL Above the Portland, first the sea becan 
or was replaced by fresh water, next followed a chain of e^ 
lagoons, extending from Hanover to the island of Ol&tin, and ' 
fresh-water lakes made their appearance over the area between 
and the west coast of Portugal. The sea, which at first had i 
retired to the south, now, at the time of the lower Neocomiau, 
over the north of France aa f ar as the Cher ; in Spain it only a 
Valencia, and in Portugal passed over scHue districts in the wes 
Hils (Hauterivian) and the Utgonian, the strand-line continually 
by oscillations until the whole region of the fresh-water lakes wt 
covered by the sea. In many places the fresh-water beds, and in 
the gypsom also, strike out against the Atlantic Ocean, proving tl 
epoch continental land extended beyond the existing coasts. 

Is temun irealdieo in nord de I'EBpagne, Bnll. Soc. g^oL de Fr., 18S6, 1 
pp. 405-M7. 

' L. Carai, itoAe des teirains ciHAcia et tertiairea du nord do I'Bipagne, 
1881. p. 147 et aeq. 

* D. Sharpe, On the Secondaiy District of Portugal which lies on the n 
Tigni, Qnart. Jonrn. GeoL Boc, ISSfi, TI, pp. 134-201 (Sub-cifitaceotu Serie*; 
Delgado, Carta geologica de Portogal, foL, 1876 ; Saemann in Loriol et Ft 
portlasdien de Boulogne, p. 184; C. Ribeiro, Breve noticia acerca da t 
lAynca e geoliigica da parte de Portogal comprensa entre ob mllei do T^o 
Jon. kI math. Ac R, Liaboo, 1870, II, pp. 243-252 and 353-369, in particnl 

* Cho&t, in Act 8oc helv. Bci. nat. r^uie a Locle 1885, 8vo, Neufch&tel, ] 



286 MESOZOIC SEAS [partih 

5. Further transgremonB cmd intermixture of the Cretaceous fauTias. 
It will be seen from the foregoing that the maximum of the negative phaae 
must be placed at the base of the gypsmn and salt marls, between the 
Portland and the Purbeck, and from this time, in spite of all osdllations, 
the preponderance of the positive movement may be recognized up to the 
transgression of the Urgonian, which covers the whole of the Weald. 

The greatest change in physical conditions did not therefore exactly 
coincide with the greatest change of the fauna, but slightly anticipated 
it, for the upper Purbeck contains a fauna which is still undoubtedly 
Jurassic. 

First of all the question arises whether this negative phase may not 
find in some other part of the world a complementary positive phase. 
Indications of this are not far to seek. The Eammeridge of the western 
region extends, as we have seen, through Poland to the Volga and beyond 
it. But even in west Poland, near Tomaszow, on the Pilica, Mikhalflki hjfts 
observed the vestiges of a sea alien to western Europe, viz. the lower Volga 
Btage of Russian geologists (zone of Periaphinctee virgatus). Mikhalski has 
described it, and doubtless with truth, as the first indication of a great 
northern infra-Cretaceous transgression \ Nikitin places the Volga stage 
in the period of the English Weald 

Since a considerable number of species characteristic of the Portland of 
western Europe still occur in east Galicia, the separation of Russia from 
western Europe must be placed at the close of the Portland, i. e. precisely 
at that time when the deposition of the gypsiferous marls begins, or at the 
maximum of the negative phase in western Europe. While in the west 
and south the strand was in retreat, leaving behind the gypsiferous lagoons, 
in the north-east on the contrary it was advancing, on the one hand over 
Kostroma, Twer, Moscow, and Riasan, and on the other from Orenburg and 
Samara over Simbirsk. 

The investigations of Russian geologists, who of late years have brought 
to light so many new facts in connexion with this subject, hardly permit 
us at present to trace in detail on the map the distribution of the several 
subdivisions of the Volga stage, but there appears to be no doubt that in 
the north of Asia a great transgression corresponds to this stage, particularly 
to its upper members. The area afiected by this transgression includes, 
according to Nikitin, the region west of the lower Ob southwards as far as 
lat. 63"" N., then a part of north Siberia, particularly the peninsula of 
Taimyr, down to Dudinskoj on the lower Yenisei, then down the lower 
course of all the rivers situated to the east as far as the Jana, but not so 
far south as the Arctic circle. There the deposits of the Volga rest in 
patches on the marine Tria& The Jurassic beds said to occur in regions 

^ A. Mikhalski, Notes sur les coaches k Perisphinetes virgatus de la Pologne ; BolL Com. 
g^l. Russie, 1886, V, pp. 363, 456. 



OH. VI] VOLGA AND UITENHAGE STAGES 287 

■ 

more to the south are more or less doubtful ; the upper Volga beds, on the 
other hand, appear to occur on the Amur and the Bureja. 

The Mesozoic deposits of Spitzbergen, which rest on the Trias, as well 
as those of east Greenland, have already been assigned, on the evidence of 
existing observations, to the Jurassic (II, p. 69); it must however be 
added that Lahusen,- in his latest investigations of the upper Volga stage 
on the Y^iisei and Lena, has observed much similarity between these beds 
and those of Spitzbergen and east Greenland ^. 

The uncertainty as to the stratigraphical position of certain beds 
characterized only by Aucella, and the difference of opinion existing among 
palaeontologists on this point, render it impossible to extend these compari- 
sons to the Aleutian islands or the Queen Charlotte islands, where the 
Volga stage is certainly represented. According to White it is very 
probable that the Aucella beds of California correspond with those of the 
north ; possibly they extend to New Mexico \ In Arctic North America, 
however, and as far as cape Farewell, nothing similar is known. In west 
Greenland the lower Cretaceous is only represented by plant-bearing beds. 

The genus Aucella can be traced from the north to the Lofoten islands, 
but it does not proceed further into the Atlantic. 

Some indications of similar relations may be observed in the southern 
hemisphere. 

In east Australia there is evidence of a Cretaceous transgression, which 
Nexmiayr places at about the horizon of the Aptian (II, p. 155). In the 
south-east of Cape Colony plant-bearing deposits, which include marine 
beds, the well-known Uitenhage series, rest on the older stratified formations 
(I, p. 899). Near Sripermatur, on the east coast of Hindustan, some sand- 
stones occur, which may be assigned to the Uitenhage horizon (I, p. 409). We 
now reach the Jurassic region of Cutch. The various stages of the 
European Jurassic may be distinguished up to the horizon of the Portland. 
The Uihia group, which contains species common to the European Portland, 
is a sandstone 8,000 feet thick. The greater number of marine fossils, 
chiefly Portland species, occur in a calcareous conglomerate near its base ; 
higher up, many terrestrial plants appear, and above the plants we meet 
with a few species of the Uitenhage stage. Here, therefore, the European 
Jurassic system is covered by the Uitenhage, a stage unknown to Europe ; 
just as in Russia it is covered by the lower Volga stage, also with an alien 
fauna. The Uitenhage fauna has come from the south, the Volgian from 
the north. To complete the correspondence, we find in Cutch, above the 

' J. Lahusen, Die Inoceramen-Schichten an dem Olenek und der Lena, M^m. Acad. Imp« 
Sci« Saint-P^tenb., 1868, XXXIII, No. 7. AmaUheus Nathorsti, Lundgr., of Spitobergea 
is also said to occur in the upper Volga stage on the Bureja. 

' C. A. White, On the Mesozoic and Cenozoic Palaeontology of California ; Bull. U.S. 
Geol. Surv. Territories, No. 15, 1885, p. 22 et seq. 



288 MESOZOIC SEAS [PAEEin 

Uitenhage stage, the Aptian, with its European fauna (Hoplitea Deshayeaii), 
and the same stage (also containing H. Deshaye^ii) rests in Simbink, 
according to Nikitin, on the Volga stage. A species belonging to tiie 
Aptian stage of Australia occurs with the European Aptian in Cutch. 

We have then in Simbirsk, first the European Jurassic, next the northern 
Volga stage, and then the European Aptian stage; and similarly in Catch, 
first the European Jurassic, then traces of the southern Uitenhage stage, 
and last traces of the European Aptian stage. 

We have now advanced far enough to return to Europe, and there to 
trace the events which mark the limit between the Jurassic and Cretaceous 
systems, as well as certain phenomena in the succession and composition of 
the marine Cretaceous fauna. 

The marine Jurassic fauna, so far as we can trace its history, did not 
give rise to the marine fauna which succeeds it. Some of the pelagic forms 
of the Alpine area may have contributed direct descendants, but the fauna 
of moderate depths was destroyed, first by the extensive retreat of the sea, 
and next by fresh-water conditions ; some brackish-water species may have 
been derived from marine forms, and have persisted for awhile ; bat the 
new marine faunas came, for the greater part, from elsewhere. 

The Valengian of the lower Neocomian, which extends but a slight 
distance beyond the pelagic region of the Alps, does not contain a rich 
fauna, and shows foreign affinities. 

The next stage is the Hils or Hauterivian, which possesses a much 
wider distribution. When Neumayr and Uhlig described the Ammonites 
of the Hils in 1881, they showed much more clearly than had been done 
before the presence of northern elements in this member of the Cretaoeoos 
formation. Subsequent investigators, among others Weerth and Marie 
Pavlow, have shown that relations exist between the Hils fauna and fonns 
occurring in Bussia ; indeed some species of the Volga fauna appear in the 
Hils. But Neumayr and Uhlig have shown that a close resemblance also 
exists between some forms of the Hils and the* Uitenhage fauna. The 
period of transgression was also a period of intermixture of faunas, and 
both northern and southern elements make their appearance in Europe. 
The isolation of the north-east^ by the negative movement, during the 
continuance of which the development of the north Volga fauna was in 
progress, now came to an end \ 

^ M. Neumayr and V. Uhlig, Ueber Ammonitiden aus den Hilsbildangen Nord- 
Deutschlands, Palftontographica, 1881, XXVII, p. 74 ; 0. Weerth, Die Fauna des 
Neocomsandsteins im Teutoburger Walde, PalAont. Abh., edited by Damei and 
Kajser, 1884, II ; Marie Pavlow, Les Ammonites da Groupe OUostopkanuB temedor. 
Bull. Soc nat. Moacou, 1886, LXII, pp. 27-48. Neumayr had inferred the noithem 
transgresuon even much earlier, not only from the marine faunas, but also from 
Sandb6tger*s statement that in the upper Weald the tropical lacustrine &una is replaced 
bj one of North American character; Verb. k. k. geol. Reichs., 1878, p. 290. 



CH.VI] SOUTH AMERICAN TYPES IN EUROPE 28^ 

Hitherto but little has been said of the lower Cretaceous deposits of 
the Alps, which succeed each other in unbroken pelagic development, 
becaufie indeed general transgressions or regressions can seldom be demon- 
strated in folded mountains ; the case of the Purbeck in the Jura mountains 
worked out by Maillard is a rare exception. On the other hand, we must 
not n^lect to take into account the character of the faunas. If we look 
for the pelagic representatives of the littoral deposits we find, immediately 
above the equivalents of the EQls or Hauterivian, the Barr^ian stage in 
the lower Rhone valley, and the Wemadorf shales in the Carpathians, and 
again a new element appears in their fauna. The rich lower Cretaceous 
fauna of Santa F^ de Bogota in Columbia, described in detail by Earsten, 
sends numerous representatives into this stage. Coquand and d'Orbigny 
recc^nized them in the BarrSmian of France, Hohenegger in the Car- 
pathians ; Uhlig has shown the correspondence which exists between the 
far-distant faunas of Bogota, the lower Rhone valley, and the Carpathians. 
The genus Pulchellia is as typical of this immigration into Europe, as 
Holooetephanus of the preceding transgression \ 

In the Aptian stage finally a uniform flora appears to have extended 
from Russia across Europe, and the few species so far known from India 
indicate that they too once extended over this region. 

Thus towards the close of the Jurassic period Europe once more 
emerges from the sea, to again disappear beneath successive advancing 
transgressions. The Alpine region remained constantly submerged. The 
mayimum of phjrsical change, between the Portland and Purbeck, did not 
coincide, in central Europe, wij^h the maximum change in the fauna. Its 
chief efiTect was to produce increased isolation and to prepare the way for 
the extinction of the sublittoral elements of the fauna. The effects of the 
succeeding transgressions on the faunas vary in importance in proportion 
as, in each case, they establish new marine communications, and thereby 
faolitate or promote fresh immigrations. 

Nikitin has pointed out that in Russia the highest degree of indepen- 
dence occurred at the time of the Volga stage, but afterwards in the 
epoch of the Gault the universality of the fauna was * gradually 
re-established even in the most distant regions.' Neumayr has arrived at 
similar results ^ Phenomena of this kind cannot be explained by oscillations 
of ihe continents. 

The distinction between the Pacific and Atlantic regions now appears 
still more sharply than before. Lower Cretaceous beds are known at many 
locaUtieB on the east coast of America, from the Aleutian islands to cape 

' y. Uhlig, Die Cephalopodenfaima der Wernsdorfer Schichten ; Denkschr. k. Akad. 
WiflB. Wien, 1883, XLYI, p. 1^. 

* Nikiiiii, torn, dt., p. 289 ; Neumajr in nimierons publications, in particular 
BidgoBchichte, II, 1887, p. 866, and Neuea Jahrb. Min., 1887, II, p. 279. 



290 MESOZOIC SEAS [PAKrm 

Horn ; also in New Zealand and Australia. On the border of the Fadfie 
they are folded; in the interior of Australia they lie flat. But on the 
Atlantic coasts, except in the case of Europe, we seek in vain for repre- 
sentatives of any of these deposits. Even in Europe the Weald advances 
with its fresh- water deposits right up to the brink of the Ocean. Never- 
theless a communication of the lower Cretaceous seas of Europe with those 
of the west of South America is indicated by their fauna. This com- 
munication, however, can only have existed in the vicinity of those parts 
of the coast which are constructed on the Pacific type, L e. between the 
existing Mediterranean and the West Indies. 

6. The Cencmuinicm transgreasion. As the long series of intermittent 
positive movements from the Rhaetic stage on through the lias was only 
the forerunner, as it were, of the great transgression of the Bathonian stage 
and the Eelloway, so. all the oscillations of the lower Cretaceous seas in 
Europe yet mentioned were only the preliminary warnings of a transgres- 
sion of far greater importance, especially distinguished by the fact that 
U affected the Atlantic coasta also. 

This transgression embraces the Cenomanian, Turonian, and Senonian ; 
there are indications that its greatest extension coincided with the Senoniaa 
Since this subject has been discussed elsewhere we will now only give 
a summary of the most important results \ 

The Cenomanian transgression has left its traces on the summit of the 
Spanish meseta; thence it may be followed to the north, across France, and 
into the north of Scotland, where its deposits may be seen at sev^ 
localities. Isolated patches of the Senonian have been revealed by dredging 
off the Norwegian coast, up to the highest latitudes (II, p. 67). The 
transgressive Cretaceous is to be seen on the summit of the Bohemian 
mass, as far as the Bavarian Jura, and further north in numerous patdlies 
in Denmark and as far as Scania. Signs of the transgression also occur in 
A part of the Baltic It extended across Poland, and Earpinsky, who has 
^<2learly recognized the difference in extension of the lower Cretaceous and 
this new transgression, describes its northern limit as having passed some- 
what north of Vilna and Mohilew through Elaluga towards Moscow, and 
then somewhat north of Simbirsk and Samara towards Orenbexg^ Thus 
the northern half of Bussia remained uncovered by the sea ; the soutbem 
outrunners of the Ural on the other hand were submerged, but Herr Kar- 
pinsky has kindly informed me of the remarkable discovery, made by 
Herr Fedoroff, that Cretaceous beds with Baculites exist beyond the Uial 
on the northern Sosswa in lai 62'' 30^ N. These Baculite beds rest upon 
fossiliferous strata, probably belonging to the upper Volga stage, and the 
whole succession is horizontaL 

' Entstehung der Alpen, pp. 104-117. 

' Earpinsky, Uebersicht der phyBiko-geogiaphlBchen Verh&ltnisse, etc, fig. 9. 



-I ipi,ui muK I , ^ .jw'wi!^ ' » 'A |> 



f 



CH.VI] THE CENOMANIAN TRANSGRESSION 291 

The Cenomanian sea extended over the whole region of the Caspian 
and the Ural ; and according to Bomanowski and Musketow it covered 
the Ej2dl-Kum and all the plains of Turania up to the great mountain 
ehain, which no sea appears to have reached since the Palaeozoic aera ; for 
the lower members of the Mesozoic series are represented in it either by 
coaI*bearing beds or not at all ^ The Cenomanian sea continued further 
into the basin of the Tarym, on the south side of which Stoliczka en- 
countered, to his great astonishment, evidence of the Cretaceous trans- 
gression. 

The middle and upper Cretaceous also cover the whole region of the 
central Mediterranean and even extend beyond it. Represented chiefly 
by hard limestone, they pass from the south of Europe across Syria, cover 
4Jbe whole eastern half of the Sahara and Arabia, and penetrate into the 
valley of the Narbada, thus reaching India. 

Marine intercalations of the Senonian are found in the plant-bearing 
beds of Disco in lat. 70" N. Patches of middle Cretaceous also occur on the 
east coast of the United States, as in New Jersey, for example. From the 
gulf of Mexico the transgression extended into the country to the west 
of the sierra Madre ; it reached Chihuahua, Praesidio del Norte, and the 
llano Estacado of Texas, and extended into the prairies through Kansas, 
Nebraska, and Dakota into the basins of the Saskatchewan and the 
Mackenzie up to lat. GS"" N. and perhaps even to the Arctic Ocean. 

At the same time the Cenomanian sea entered the valley of the Amazon, 
crossed the whole breadth of South America to the Andes, and probably 
reached the Pacific Ocean. Its sediments are seen at Piauhy, Ceara, and 
as far as Bahia (I, p. 510). In the south the Cretaceous forms the under- 
lying rock of the Pampas, and middle Cretaceous fossils are found up to 
iat. 58** 30" S. in Patagonia (I, p. 526). 

On the west coast of Africa from near the equator to Mossamides 
{n, p. 184) deposits of the lower Cenomanian occur, which some observers 
refer to the Qault, but on the east coast they reappear with a different 
fauna^ which has been distinguished as the Indian type. This fauna 
accompanies the transgression on the coast of Natal, again on the coast 
of Trichinopoli, and finally in the Shillong plateau (I, p. 411). West 
Australia is covered by transgressive middle Cretaceous beds, concerning 
which, however, we unfortunately possess but scanty information. 

The Pacific region has suffered so much from compression and folding 
that it is unusually difficult to prove the existence of transgressions within 
it. In Yezo and Saghalin the Cretaceous of Indian type reappears, 
probably in transgression (11, p. 182) ; middle Cretaceous lies unconformably 
and horizontal in the foot-hills of California (I, p. 584). 

' 6. Bomanowski, Materialien zur Geologic von Turkestan, German edition, 4to, 
Si. Petersburg, 1880, p. 48. 

u a 



292 MESOZOIC SEAS [PABrm 

ThuB the sea of the latter half of the Cretaceous period covered Torania 
and Irania, Arabia and the Sahara, the prairie-land of North America and 
the great valleys of the Amazon and Parana. On the other hand a definite 
area may be recognized which does not appear to have been reached by 
the inundation. This embraces Greenland, Spitzbergen, perhaps the north 
of Scandinavia, northern Russia, Siberia, and the whole of northern China. 
The conjecture might have been hazarded that the Cretaceous formation 
had been removed from this region by ice, but the Tertiary leaf -bearing 
beds are everywhere present, and rest on the Jurassic or Volga stage in 
east Greenland and Spitzbergen, while the middle or upper Cretaceous is 
not to be seen ; it is true, however, the Baculite beds of the Sosswa lie 
within this region. A great part of Gondw&na land, broken up even so 
early as this period, also remained above water, for we see in the east u 
in the west of Africa, that the Cretaceous beds only rest against the sides 
of the plateau, or, as in India, at most extend into the valley of the Narbada, 
without reaching the plateau itself. West Australia, on the other hand, was 
submerged. 

However this may be, it would appear, according to the present state 
of our knowledge, that a considerable area of dry land extended around the 
north pole, particularly on the Asiatic side, and more especially in the 
direction of northern China. In the lower Cretaceous period the Volga 
stage came from the north southwards, the Uitenhage stage from the south 
northwards, both from polar regions. The middle and upper Cretaceous 
transgressions, on the other hand, chiefly affected the equatorial and 
temperate zones; it is probable that the sediments in the valley of the 
Mackenzie, the marine intercalations between the plant-bearing beds of 
Disco, and the deposits in the deep water off the coast of Norway represent 
outrunning arms of the sea. Whether this explanation will apply to the 
beds exposed on the Sosswa, future investigation must decide. 

The great extension of the middle and upper Cretaceous seas was followed 
after the Senonian, as it was after the Portlandian, by an extraordinary 
retreat. This negative phase will be discussed in connexion with the Eocene. 

7. Survey of the Meaozoic seas. The transgressions covering the con- 
tinents of the Mesozoic period, whether they came from the poles or the 
equator, nowhere left sediments on the continents which would indicate 
abyssal depths. The Eelloway, the transgressive beds of the lower 
Cretaceous, and the Cenomanian are composed of clastic materials, ac- 
companied as a rule by numerous organic remains, and it is only in the 
Senonian that a sea of any considerable depth can be supposed to have 
covered northern Europe. The Callovian of Russia and the Cretaceous 
mantle of the upper valley of the Elbe may be cited as examples. These 
mark merely temporary submergence of the land, and must not be ooufiised 
with the great persistent abysses of the Ocean. 



^.■•■4^« 



CH.VI] REVIEW OF THE MESOZOlC SEAS 298 

Onr geographical maps show the outlines of existing seas, but some of 
these are very deep ; others, on the contrary, as for instance the Arctic 
Ocean, are of very trifling deptL A knowledge of the outlines does not 
imply therefore a knowledge of the position of the great depths. 

It is clear that the Trias deposits around the Pacific Ocean are repre- 
sented in the folded chains ; in the Indian Ocean the evidence of a trans- 
gressive border or overlap only begins with the middle Jurassic, and in 
the Atlantic region with the middle Cretaceous. In the Indian and 
Atlantic Oceans we observe besides that the border is not folded. 

In that part o| the Arctic Ocean which washes the coasts of Asia as 
far as the east coast of Greenland, and the Lofoten islands, but not beyond, 
certain features appear in common with the Pacific region. As far as 
Spitsbergen the Trias is similarly present, but in Spitzbeigen it is not 
folded. Certain parts of the Jurassic system as well as the Volga stage 
appear in these parts of the Arctic regions, but the middle and upper 
Cretaceous are absent. Baffin bay, on the other hand, shares the characters 
of the Atlantic Ocean: the only marine deposit of western Greenland 
belongs to the upper Cretaceoua 

We will now leave the shore of the Ocean and cast a glance at the 
stratified series of the Silakank in the Himaklya (I, p. 487). Where, as in 
this case, the Mesozoic beds succeed each other uninterruptedly in per- 
sistent pelagic development, for a thickness of thousands of feet, and even 
overlie a similar series of Palaeozoic beds, we seem forced to conclude that 
a deep and long persistent sea lay in the midst of the existing continent, 
and that it was brought to an end by the accumulation of- its own sedi- 
ments^ by negative movements of the strand, and by the folding of the 
mountains. The eastern Alps afibrd us examples which are no less striking, 
and we must admit with Neumayr the existence of a central Mediterranean, 
which extended as early as the Trias from Asia across the south of £urope, 
over part of the region of the existing Mediterranean to beyond Gibraltar. 
That this sea passed across the Atlantic even before the deposition of 'the 
Cenomanian on the coasts around this Ocean is proved by indications in 
the Jura, and by characters even more definite in the lower Cretaceous 
marine fauna of Bogota. 

The enlargement of this central Mediterranean, which originally ex- 
tended parallel with the equator, may have subsequently given birth to the 
Atlantic by a succession of subsidences. Even in the Weald the fresh- water 
beds on the coast of Portugal and north Spain are cut off abruptly by the 
sea. The occurrence in the West Indian islands of many corals charac- 
teristic of the Turonian of Europe presupposes a coast-line, or at least 
a number of fairly large islands in the midst of the existing Ocean, to 
render the migration of the corals possible (I, p. 281). 

It is, therefore, probable that even before the time of the Cenomanian 




294 • MESOZOIC SEAS [PAirrra 

the ancient sea which stretched from Europe to the West Indies had 
experienced a considerable extension by the addition of fresh subsidences, 
and that from this extended region the Cenomanian transgressions pro- 
ceeded, but at the same time it is equally probable that since then an 
increase has taken place by the occurrence of yet other subsidences. 

The mutual encroachment of transgressions and tectonic processes 
renders a complete solution extremely difficult, and we are compelled to 
express ourselves with great reserve. Steinmann met with Cretaceoiis> 
sandstone in Bolivia, lying in horissontal beds at a height of 4,000 metres 
above the sea. He was led to conclude that since its formation 'the 
surface of the sea must have moved so much nearer to the centre of the 
earth \' 

For the present we may content ourselves with the conclusion that the 
existing oceans are of different ages, and turn our attention to another 
question. 

In the opening pages of this book the outlines of the continents, nar- 
rowing away to the south, were described as the most striking feature on 
the map of the world ; and it was further observed (I, p. 5) that any 
attempt to comprehend the movements and changes of form of the earth's 
crust must take into account these great features of the planetary surfoee. 

The wedge-shaped form appears in four typical cases: in Soath 
America, South Africa, India and Greenland, and thus in the most diverse 
latitudes. Other examples of less importance might be cited, such as the 
peninsula of Sinai and the Crimea ; even in the interior of the continents 
similar outlines may be seen, as for example in the Bohemian mass. 

South America, formed in the west and south by the extremily of 
a curved mountain range, in the east of alluvial land, deviates completely 
from the other examples in structure, and may be left out of account The 
three remaining examples are table-lands. Greenland is a radial slice of 
the Atlantic, the table-land of Old Bed sandstone: India is a similar 
segment of the GondwAna table-land : South Africa is another part of the 
same table-land. 

Greenland exhibits beds of the Asiatic Arctic type resting against its 
eastern coast, and beds of the Atlantic type on its western coast The 
coast-lines of different seas meet in cape Farewell. These facts cannot he 
explained by the elevation theory. Greenland is a horst of the first order 
between two or more sunken areas of different age. 

India rises in the same way between the bay of Bengal and the Arabiaxi 
sea, and South Africa between the two Oceans : in these cases the distriba- 
tion of the Gondwina formations and the lie of the thick plant-bearing 
beds, which in the case of Africa boldly face the sea, are sufficient proof 

^ Steinmann, Gompte rendu, 69°^® Seerion de la Soci^t^ heW^tienne; Arch. scLphjfc 
et nat. Geneve, 1886, p. 93. 



CH.VI] WEDGE^HAPED OUTLINES 295 

of the fracture. The various deposits resting against these masses indicate 
the age of the different coast-lines (I, p. 418). The elevation theory cannot 
explain how a table*Iand of plant-bearing beds, such as this, which has 
never been covered by marine deposits, could have been raised from the 
depths of the sea. 

In like manner the peninsula of Sinai, the Crimea, and the Bohemian 
mass are bounded on the south by the convergent edges of regions of 
sabsidence. 

Thus, as our knowledge becomes more exact, the less are we able to 
entertain those theories which are generally offered in explanation of the 
repeated inundation and emergence of the continents. 



296 



CHAPTER VII 

TERTIARY SEAS AND RECENT LIMESTONE 

FORMATIONS 

Negative phase at the close of the Cretaceous period. The central Mediterranean of 
the Tertiary aera. The east coast of North America. The Tertiaiy region of Patagonia. 
Recent limestone formations. Summary. 

1. Negative phase at the close of the Cretaceoue period. Towards the 
end of the Cretaceous period events occurred similar to those which char- 
acterized the later ages of the Jurassic period. The sea was reduced in 
area. The prairie land of North America, from Canada down to Texas and 
Alabama, was laid dry, and the Laramie stage, formed of alternating brackish 
and fresh-water beds, corresponds to this period of the retreating strand. 
The mammalian fauna of the Tertiary aera had not yet made its appear- 
ance, but great reptiles were the most striking representatives of the 
terrestrial fauna, as in the Cretaceous period. During the same epoch 
the basin of the Amazon was abandoned by the sea : the brackish-water 
deposits of Pebas are the only traces so far known of the influence of the 
sea in post-Cretaceous times; they belong to the Eocene or Oligocene 
period (I, p. 512). Simultaneously the strand receded from north to south 
in England, the horsts of central Europe were abandoned by the sea; in 
Russia the shore receded to the south. Far and wide fresh land made its 
appearance. 

When the negative phase, towards the end of the Jurassic period, had 
attained its maximum, the sea occupied the central Mediterranean, in other 
words the site of the younger folded ranges of western Eurasia ; but outside 
this region, over an area extending up to its borders, as in the Jura for 
example, deposits of gypsum were first laid down, and then an alternation 
of fresh- water and marine beds which still bore a Jurassic f acies. 

Events followed much the same course towards the end of the 
Cretaceous period. 

From Spain through the south of France, and particularly in Provence, 
we find, resting on the marine Cretaceous, fresh-water beds, which are 
followed in certain areas by a series with marine shells of Cretaceous type. 
^liis is Leymerie's Oarv/mnian stage. It occupies the same position with 
regard to the Cretaceous system, as the Purbeck to the Jurassic. 



CH.vn] GAEUMNIAN STAGE 297 

The Garumnian stage has been stndied in Catalonia by Vidal, by 
Hallada in Huesca, and by Leymerie, Matheron, and others in the south of 
France ^ 

In the Jura we found it possible, notwithstanding the folding, to deter- 
mine the outlines of the several lacustrine areas, which were formed at the 
close of the Jurassic and the beginning of the Cretaceous period, and 
similarly in the present instance, in spite of foldings and various sub- 
sequent dislocations, we are already in possession of data, thanks to the 
zeal of observers, which will enable us to ascertain the course of events 
during this phase, equally negative, which marks the transition from the 
Cretaceous to the Tertiary aera. 

The lowest member, as in the Jurassic, ia the most restricted in dis- 
tribution. It consists of the lignitiferous fresh-water beds of Fuveau. 
These rest on the upper Cretaceous, which has already acquired a brackish 
water character: they occur only in the neighbourhood of Marseilles, 
extending westwards from this town as far as Martigues on the £tang de 
Berre, and for about the same distance to the north and east. They were 
probably deposited at the mouth of a river. 

Above them come the fresh-water limestones of Rognac, distinguished 
by Lychnus and other fresh- water genera, which do not pass upwards into 
the beds above: the faima, while purely fresh-water, is still thoroughly 
Cretaceous; land reptiles of Cretaceous age are also present. Over wide 
areas this division begins with a bed of bauxite (aluminium hydrate). It 
extends from the Yar across the Bouches-du-Bhdne, Vaucluse, Gard, 
H^ult, and Aude, and passes, in Ari^e, into continually closer relations 
with the uppermost beds of the marine Cretaceous, which, in the Haute- 
Garonne, loses more and more its fresh- water intercalations ^. The Bognac 
beds also occur in the north of Spain. 

The third member consists in the east of layers of red clay, sandstone 
and pebble beds ; to the west, on leaving the Rhone, the red clays increase 
in thickness; they are known as the ' argUes nUUarvbes,' their brilliant 
colouring rendering them a conspicuous feature in the landscape; they 
extend over Hdrault and Aude, and reappear with the same characters 
beyond the Pyrenees, still possessing a considerable thickness. They are 
very poor in f ossila In Provence the upper beds contain species of Physa, 
Lymnaea, and Planorbis, existing genera of pond and marsh snails ; these 
continue into the succeeding beds. All Cretaceous types have now dis- 
appeared, among them the genus Pyrgulifera, which however still main- 

' We need only mention here the succinct account by P. Matheron, Note rar les 
d^pdts cr^tac^s lacustres et d*eau saum&tre du Midi de la France ; BulL Soc g^L deFr., 
1875-1876, 3« s^r., IV, pp. 415-428. 

' CroisierB de Lacvivier, Etudes g^ologiques sur le d^partement de TAri^ge ; Ann. sci. 
gM., 1884, XV, pp. 1-293, in particular p. 250. 



298 TEKTIARY SEAS [pakthi 

tains its existence in lake Tanganyika. Above the days follows first 
fresh- water limestone, with Physa, and then the marine Eocene. 

It is thus the red days which mark the division between the faunas. 
At their base the rich <>etaceoTis fauna of Bognac has disappeared, and 
from this time onwards the younger forms prevail. Matheron has brought 
this out in clear relief, and Boule, in his admirable description of the ex- 
posures in Provence, has placed the limit between Cretaceous and Tertiary 
at the base of these clays \ 

The same interruption of the marine series between the Cretaceous 
and Eocene is repeated in the region north and east of the Adriatic; 
Stache's investigations show that from Carinthia, through Istria, the 
islands of the Quamero and a great part of Dalmatia, a varied group of 
intercalated brackish or fresh- water beds appears on the horizon. Near 
Sebenico the fresh- water limestone rests on an eroded surface of Cretaceous 
limestone. A pre-Tertiary red earth also takes part in these formationa 
The whole group of beds has been named by Stache the Lihv/mian stage, 
and correlated, at least in its lower part, with the Grarumnian stage K 

Thus at the dose of the Cretaceous period the sea had again beotme 
extremely restricted. We now reach the first great stage of the marine 
Tertiary deposits. As we approach nearer to existing times the number of 
recorded observations increases, but at the same time the immense variety 
of facts becomes almost overwhelming, and description correspondingly 
difficult, since so much has to be condensed. To give in a few pages 
a fairly adequate account of the subject I must refndn almost completely 
from citing the sources of my information, and I must restrict myself 
to discussing the Tertiary and more recent marine deposits in three 
regions only. 

The first of these is that which we have already alluded to as the 
central Mediterranean ; the second is that of the west Atlantic coast from 
lat. 48"" N. to the valley of the Orinoco ; the third is formed by the Fata- 
gonian plains from Paran& to Tierra del Fuego. 

With regard to Australia, especially the Bunda plateau, and the absence 
of Tertiary deposits on the east coast, I must refer to what has been said 
already (11, pp. 162, 162). The Tertiary deposits of the Pacific coast ate 
unfortunatdy still little known. The T^jon group of California, which has 
recently been assigned to the Eocene, contains Ammonites, and differs in 
several respects from the European and the eastern American types. The 
same is true of the deposits of the island of Quiriquina, in Chili, which are 

' L. Roule, Recherches sur le terrain fluvio-lacostre infSrieur de la Provence ; Ann. ml 
g6ol., 1885, XVIII, in particular p. 129. 

* G. Stache, Die libnrnische Stufe, Verh. k. k. geol. Beichs., 1880, pp. 195-209 ; bj the 
same, Ueber das Alter von bohnerzfohrenden Abla^mngen am ' Monte Promina * in 
Dalmatien, op. cit., 1886, pp. 885-887 et passim. 



CH.vn] THE CENTRAL MEDITERRANEAN 299 

referred by some authors to the Cretaceous, by others to the Tertiary^. 
Folded Tertiary beds occur in the coastal chains of California, but their 
exact age has yet to be determined. 

2. The central Mediterranean of the Tertiary aera. By the region of 
the central Mediterranean we understand, according to Neumayr's definition, 
a broad zone embracing the Betic cordillera, the whole Alpine system, and 
the larger part of the great ranges of Asia ; from the Trias upwards it is 
distinguished by an uninterrupted series of marine sediments. When 
towards the end of the Jurassic, a negative phase set in over a considerable 
part of Europe, and the Purbeck and Wealden beds were deposited, this 
r^on still remained covered by the sea. The same was the case when the 
negative phase, which occurred at the limit between the Cretaceous and 
Tertiary, made itself felt. In the central Mediterranean region marine 
Cretaceous beds are covered by marine Tertiary beds, and both have sub« 
sequently been folded. But this continuity is not confined to great folded 
ranges ; Zittel has shown that the marine Eocene beds of the Libyan desert 
rest directly upon the marine Cretaceous ^. The Garumnian stage in Spain 
and the south of France, as well as the libumian stage in the region of the 
Adriatic, only represent marginal deposits of this sea, which doubtiess was 
very much restricted. 

The strand having sunk to its lowest level now began to ascend ; 
through many oscillations the Eocene sea extended itself over the basin of 
Paris and a part of Belgium into the south-east of England. Strange and 
remarkable faunas appear and disappear, like those of the sands of Bracheux 
and the limestone of Mons, until finally the typical beds of the French 
Eocene, the Sables inf^rieurs and the Calcaire grossier of the Paris basin, 
were laid down. It is also evident that the Eocene sea, extending from 
the Atlantic, invaded the valley of the lower Loire. At the same time its 
sediments were spread out from the Carpathians and the Crimea over a large 
part of south Russia. 

From the Betic Cordillera the Eocene sea extended to north Africa ; it 
embraced a large part of the eastern Sahara, then Syria, Arabia and Persia. 
From the great ranges of India it spread over Cutch and Ouzerat, and 
over the plateau of Shillong. 

Its sediments occur everywhere in the folded mountain chains, pro- 
ceeding from the west through the Alps and Carpathians, the Apennines, 
and through the Crimea to the EQm&laya: between the inner chains of 
the Himalaya a band of marine lower Tertiary has been traced for a 

* C. A White, The Chico-T^jon Series, Biill. U.S. Geol. Surv. Territories, 1885, No. 15, 
pp. 11-17; R. A. Philippi, Ueber die VerBteineruogen der Terti&rformation Chiles, 
Zeiticlir. gesammt. Nat., 1878, 3. Folge, III, pp. 674-684 ; near Qairiquina Plesiosaurus, 
Baenlites, Trigonia occur in tiiese beds. . 

* JL Zittel, BeitrOge zur Geologic und Pal&ontologie der Libyschen Wttste, 4to, Cassel, 
1883, p. *xc 



300 TERTIARY SEAS [PAKPin 

distance of 800 kilometers, and it rises above Leh to a height of 21,000 
feet (I, p. 488). The farther continuation of these sediments in tiie great 
ranges of Thibet is not known, but Eocene deposits are seen again in Luson. 
From the plateau of Shillong they extend into the folds of Burmah. The 
older Tertiary beds of the Malay arc from Sumatra to Borneo are likewise 
of marine origin: there is much difference of opinion as to their precise 
age. Some traces of Eocene have been said to occur in Madagascar also, 
but this statement has not yet received confirmation. 

Notwithstanding the fragmentary state of our knowledge it is evident 
that sediments of the Eocene sea extend over the south of Eurada in 
folded ranges, and that outside these ranges otrtposts of flat-lying deposits 
occur ; in the north, as far as England and over the south of Russia ; in the 
south, from the Sahara across Arabia, Cutch, Quzerat, and to the Brahma- 
putra. In breadth, this sedimentary zone, interrupted by islands, extends 
from London to E^rtoum, and from Elev to the Indian Ocean. 

NotwithstanAmg its wide extent, the aediTnejUa of this sea are nowhere 
seen to overstep the limits of the Cretaceous area* In Europe its distributioa 
is more restricted than that of the Chalk ; in the whole of the south from 
the Sahara to Ouzerat its sediments terminate in an escarpment, or, as in 
Arabia, in steep cliffs, or again, as in Shillong, by a flexure ; they are every- 
where superposed normally and conformably on the Cretaceous, and their 
southern shore is unknown. 

In the west of Europe, where more detailed observations are at our 
disposal, a negative phase may again be recognized. The south of England 
was laid dry, and the Hempstead series was deposited in fresh water. The 
valley of the Seine was also abandoned by the sea, and the gypsum ijf 
Montmartre was deposited ; beds of gypsum belonging to this period extend 
into Provence and into the neighbourhood of Miihlhausen in Alsace. Over 
northern Germany, which was not covered by the sea, a great formation of 
brown coal accumulated. This negative phase marks the boundary between 
the Eocene and Oligocene. 

Once more Europe experienced a positive movement of the strand, and 
the Oligocene sea made its appearance. The beds of Castel Qomberto in 
the southern Alps, distinguished by a rich marine fauna and in particukr 
by numerous corals, are known in many localities as far as Suez, as well 
as in Armenia. They extend into the south of France, overlie the fresh- 
water beds and gypsum of the preceding negative phase, are exposed at 
Bordeaux, then at Bennes, and, as the sands of Fontainebleau, enter the 
basin of Paris, where they extend to the south somewhat beyond tiie 
boundary of the marine Eocene : they are represented in England only 
by brackish-water deposits; finally we meet with them as the marine 
sands of Weinheim in the Rhine valley near Mainz, and even above this 
city. These marine sands are covered by blue clay, the septarian clay, or 



cH,vn] THE OLiaOCENE TRANSGRESSION 301 

the Bupelian clay of Dumont. Opinions differ as to whether the marine 
sandB entered the Rhine valley from the north or the south, but the 
northern origin of the superimposed septarian clay Lb indisputable \ The 
ooials of Castel Gomberto are rare near Bordeaux (Qaae); in the Paris 
basin and at Mainz they have as good as disappeared, but the moUuscan 
foona is still that of a warm temperate climate. In the clay, on the other 
hand, species of boreal type predominate; this marine clay extends far 
over north Germany, through Berlin to Stettin and Eonigsberg ; it is the 
same as that already mentioned as extending, according to Earpinsky, 
over the Russian plain, and on the east side of the Urals as far as lat. 68^ N. 
(I, p. 822). About this time, therefore, the sea advanced from the north 
over the region of the Obi, iiien south of the Ural to Europe, and extended 
over Germany to Belgium. 

Thus the course of the ancient marine transgressions over the Russian 
Platform has been as follows : from the south in the Callovian, from the 
north during the Volga stage, from the south in the Cenomanian, and from 
the north in the Oligocene period. 

The shores receded once again. The Aqydtanian lignite measures and 
the lower fresh-water molasse of Switzerland were deposited in central 
Enrope. The succeeding positive phase, however, again falls within the 
range of those events, which have already been discussed in our sketch 
of the history of the Mediterranean. 

The deposits of the first Mediterranean stage extend from the Azores 
and Madeira through the south of Europe, across Asia Minor and Armenia 
to Fersiaw In his latest travels, Griesbach found Tertiary beds, which may 
be the continuation of this great zone, in Ehorassan ; they contain marine 
shells which he r^ards as Miocene, and form a belt on the southern border 
of the Aralo-Caspian area, passing through Badghis, Maimeni, and even 
to beyond Taschkugan. Even across the Oxus, near Eilif , north of Balk, 
this indefatigable investigator found inclined beds of limestone rising from 
the Turkmenen steppe, and containing Ostrea, Pecten, and Polyzoa: he 
assigns them to Abich's Salt formation K 

From these facts we perceive what a vast extension the ancient 
eentral Mediterranean still possessed in the middle of the Tertiary aera, 
when a fauna was already in existence closely resembling that of the 
present Mediterranean* The shells collected in Persia show, however, that 
the communication with India, which was stUl open during the Eocene 
period, now no longer existed. 

^ A. Andieae and W. Eilian, Briefnrechsel fiber das Alter des Melanienkalkes and die 
Herkimft des Teriiftnaeeres im Rheinthale ; Mitth. Gomm. geol. Landesantersach. Ton 
Bia^Lothr., 1, 1885, pp. 72-82. 

* C. L. Griesbach, Field-notes fiom Afghanist&n, Rec Oeol. Surv. India, 1886, XIX, 
I^ 257 ; and by the same, Field-notes No. 5 to accompany a geological Sketch-map of 
Afghanist&nand Noiih-Eastem Ehorassan, op. cit» 1887, XX, p. 199. 



302 TERTIARY SEAS [PARrnr 

In Europe the northern shore ran up the Rhone valley, included a part 
of the Jura, and proceeded from the south margin of the Black forest to 
that of the Bohemian mass, and along the eastern side of the Manhart and 
the Sudetes to Silesia. 

In the East over a vast area evaporation next set in. 

This was the time of the ScJdier, the deposit of an expiring sea. The 
salt beds of the Carpathians were then formed, and probably at the same 
time the great salt beds of Persia and Turkestan. Communication with 
the Rhone valley, across the existing Jura, had by this time ceased. 

The blue marl of the Schlier is covered here and there by fresh-water 
deposits : tectonic changes occurred in central Europe : then came the 
second Mediterranean eta^e. By this time the greater part of the East 
appears to have been abandoned by the Mediterranean, at all events its 
asserted presence in Persia requires confirmation. To the north of the 
Crimea it still persisted, even as far north as the valley of the Manytsh, 
and has left behind some scattered deposits. In lower Austria and 
Hungary it entered freshly subsided areas of great size lying in the midst 
of the mountains, but Bavaria and upper Austria were no longer submerged 
by it. 

A negative phase next followed, and isolated the Sarmaiian region; the 
whole valley of the Danube, Qalicia, the south of Russia, and the last 
remains of the Aralo-Caspian area were now abandoned by the sea. The 
Mediterranean was then restricted by another negative movement to 
a space even much smaller than that it at present occupies. Its eastern 
boundary probably lay near Corsica and Sardinia. At the beginning of 
this period eroeion of the valleys occurred, as in the Rhone valley, some 
parts of western Hungary, and doubtless elsewhere : later on great inland 
seas existed, in which the Cardium beds of the Pontic 8ta{fe were deposited ; 
but the arrival of marine fishes, which ascended the rivera to spawn in the 
lakes, suggests that the negative period had already passed its dimax, 
which corresponds in fact with the erosion of the pre-Pontic valleys. 
Up to the present we know of no true marine deposits of this period in the 
Mediterranean region, and thus Neumajrr was justified in supposing that 
the strandrliTie at this tirne was prohahly situated lower than at jirreeent 
(I, p. 835). 

A gap occurs in the Mediterranean series at this point ; it marks the 
culmination of the negative movement, and it is at precisely this horizon 
that the limit is usually drawn on palaeontological grounds between the 
Miocene and Pliocene. 

It is true that a positive phase again followed, but the sea of the third 
Mediterranean stage was far from attaining its former extent; its area 
was even much more restricted than at present, although the strand stood 
somewhat higher, since the subsidences in the region of the Adriatic, on the 



cH.vn] TEETIARY BORDER OF UNITED STATES 308 

Syrian coast, and in the Aegean, had not yet occurred Next came the 
fonrth Mediterranean stage, and the temporary northern immigration; 
this time, however, not from Siberia towards central Europe, as in the 
Oligocene period, but from the Atlantic Ocean* Local subsidences followed ; 
the Pontus was annexed, and the present state of things established. 

If we disregard the tectonic incidents, that which we observe is an alter- 
nation of positive and negative phases of different value ; in each succeeding 
alternation the positive phase had less and less effect in extending the 
confines of the sea, or its area, as compared with' the preceding positive 
phase, was reduced ; until the negative maximum supervened in the period 
of erosion which preceded the Pontic lakes. 

This gradual reduction was interrupted by a temporary transgression 
of the boreal sea in the Oligocene period. 

The gain due to the slightly higher level of the existing strand, and to 
the addition of insunken areas, was by no means sufficient to balance the 
loss resulting from the decrease in amplitude of the positive movement, 
and thus the existing Mediterranean has come to represent the remains of 
an Ocean, which even at the commencement of tiie Tertiary aera stUl 
extended over central Asia. 

An excellent idea of the way in which these phases succeeded each 
other may be obtained from Iwanow's map of the government of Stavropol. 
North of Pjatigovsk there rises a boss of quartz porphyry, surrounded by 
sandstone belonging to some early stage of the Tertiary; the second 
Mediterranean stage follows on the north, extending from Qiorgiensk to 
beyond Stavropol; it is horizontally stratified and covers a considerable 
area; towards the north, Le. towards the Manytsh, it is followed in its 
turn by the Sarmatian stage, also fiat ; this lies in a synclinal, and beyond 
the Manytsh the other flank of the synclinal makes its appearance : then 
follows within this synclinal of Sarmatian, a second, of Pontic beds; in 
this lies a third of Aralo-Caspian deposits, and in this flows the river 
Manytsh ^ 

3. The east coast of North ATnerica. With regard to this region a 
number of facts of general importance have already been discussed 
(I, p. 281) : a detailed description has been given of the stratified succession 
in Antigua (11, p. 185), so that we shall be able to treat with greater 
brevity the relations of the Tertiary beds on the west coast of the 
Atlantic with those of Europe : the facts bearing on this subject are mostly 
taken from Heilprin's latest works '. 

* Iwanow, Qeologische Karte des Stawroporschen Guberniums; Gornoi Joum., 1887, 
No. 7. 

* A Heilprin, Contribation to the Tertiary Geology and Palaeontologj of the United 
States, 4to, Philadelphia, 1884, map ; for the middle Tertiary beds of North Jersey, by 
the same, Proc. Acad. Nat ScL Phil., 1886, p. 851 ; farther, by the same. Explorations on 
the West Ck>ast of Florida and in the Okeechobee Wilderness ; Wagner, Free Inst. Sci. 



304 TERTIAEY SEAS [PARrm 

The gradual retreat of the strand is much less clearly expressed on the 
Atlantic coast of Europe than on the Mediterranean, because its outline is 
broken by rias coasts and is in general much more diversified; on the 
Guadalquivir, in Portugal, and on the Oironde, however, the retreat maybe 
readily recognized. The American coast of the Ocean possesses a simpler 
structure. On the island of Martha's Vineyard (lat. 41'' 20' N.) a zone of 
marine Tertiary strata occurs, which rests regularly against the continent 
and borders the Ocean, proceeding through the peninsula of Florida and 
the Antilles to the Orinoco, that is through more than 83 degrees of 
latitude. These beds rest conformably on the upper Chalk for great 
distances, but are separated from it in the south by an * eo-lignite ' stage, 
and they are so disposed that successively younger members appear as we 
approach the sea. From this we may draw two conclusions : first, that this 
has been a littoral region since the Cretaceous period, and next that the 
coast-line has progressively receded. 

Tet each of these results requires some modification. 

Although for 83 degrees of latitude the marine beds lie as they w^^e 
deposited since the middle of the Cretaceous, yet, apart from the early 
Cretaceous fauna of Bogota and the Cretaceous corals of Jamaica, there is 
such a dose correspondence of the Oligocene corals of Castel Gomberto 
near Yicenza and of the first Mediterranean stage near Turin with those 
of the homotaxial deposits in the Antilles, that to account for the distri- 
bution of these corals we are compelled to assume the existence of a aeries 
of islands or a continuous coast-line, which as late as the first Mediterranean 
stage extended through the tropical zone. 

The strand no doubt receded progressively, yet it is none the lees 
probable that here, as in Europe, the existing level is not the lowest which 
has been experienced. The facts are as follows : — 

The Tertiary selvage runs parallel to the coast from New Jersey 
through both the Carolines and Qeorgia ; in Georgia it is nearly 260 kilo- 
meters in breadth, and its boundary bends in south Alabama gradually to 
the north, extends beyond the mouth of the Ohio into the Mississippi, and 
from there crosses the Bio Qrande above Laredo to the south-west All 
the other members, including the Orbitoides limestone, follow the north 
coast of the gulf of Mexico; there they are covered by the deposits of 
a great iidand sea, the Qrand Qulf series, which is absent on the Atkntie 
coast (I9 Fig. 37, p. 284). At the same time, however, the Oligocene and 
Miocene, consequently the younger members of the Tertiary zone, proceed 
to the south through Florida towards the Antilles. Thus Florida is, as it 
were, one-sided, since the supposed equivalents of the European Miocene 
occur only along the Atlantic coast. 

FhiL, 8vo, Philadelphia, 1887 ; W. H. Dall, Notes on the Geology of Florida, Am. Joant 
Sd., 1887, XXXIV, pp. 161-170. 



;i ! 



CHUvn] FLORIDA STAGE 305 

m 

But Heilprin has shown, in confirmation of Conrad's statements ^, that 
while these representatives of the Miocene are present all along the coast- H 

line, from the north downwards, yet they are immediately followed by 
beds containing a much younger fauna, very similar to that now existing, '| 

and described as post-Pliocene, and further that as far south as Florida no fi 

one has yet succeeded in discovering an equivalent of the European Pliocene. 
In the gulf of Mexico, about the mouth of the Mississippi, the very 
young, certainly post-Pliocene beds of the Port Hudson group, rest directly 
on the limnic beds of the Great Oulf series (lat.29^'81'^ N.). In south Florida, 
on the other hand, new members make their appearance. On the river 
Galoosahatchie (lai iG" 30 >26^ 40" N.) the Florida stage, rich in mollusca, 
appeaiB, which Heilprin is inclined to compare with the European PHocene. 
The existing sediments appear as the direct continuation of the very flatly 
bedded sediments of the preceding period, and the gradual retreat of the 
strand finds expression in the fact that the older beds were deposited in | 

deeper water, the younger in less deep* Dall, it is true, believes that here * i. 

also he has found traces of oscillations. j ] 

The younger zones of Florida are continued into the outer zone of the ' | 

Antilles. 

In this important region of North America, where many questions 
still await solution, investigation will certainly bring to light much that 
is new: at present we may record the fact that in the region of the 
Earopeaa Mediterranean a regular negative maximum occurred between 
the Miocene and Pliocene, from which we may conclude that the strand- 
line in this region then stood at a lower level than at present, and that 
on the other side of the Ocean, in about the same latitude and still a little 
farther to the south, a gap occurs between the supposed upper Pliocene 
and the post-Pliocene, which overlies it : further to the south, however, 
in Florida, this gap has not been shown to exist. 

4. The Tertiary land of Patagonia. The Tertiary beds which advance, 
south of Parana, through 20 degrees of latitude to the Atlantic coast, 
differ from those of North America in two respects: in the first place 
they include a series of terrestrial formations, with which occasional 
marine beds are intercalated; and next, these terrestrial deposits end 
against the searcoast, in such a manner as to show that the mainland 
once extended much further to the east. In the north they proceed 
a long way up the Parani, and towards the west nearly reach the eastern 
slopes of tiie Andes. The resemblance of the basin of the Parani to that 
of the Mississippi is very striking. 

These beds form the most extensive Tertiary land in the world : their 
vide distribution in a north and south direction affords an excellent 

^ T. A. Conrad, Gotalogae of the Miocene SheUs of the Atlantic Slope ; Proc. Acad. 
Nat 8ci. Phil., 1862, p. 559. 

■GLLAfll X 



I ' 



j-^ 



306 TERTIARY SEAS [PAram 

opportunity for probing the questions with which we are now occupied. 
The latitude of the strait of Magellan corresponds very nearly with that 
of Cambridge and Birmingham^ and the latitude of Parani closely enongfa 
with that of Alexandria : hence we may hope to obtain some information 
as to the nature of the events which happened in the southern hemisphere 
contemporaneously with those in corresponding latitudes of Europe. 

Our knowledge of this subject is chiefly due to d'Orbigny and Darwin, 
Burmeister, Ameghino, and Doering, particularly to Doering, whose 
account I am now about to follow^. 

The Tertiary beds of Patagonia present no signs of dislocation. Bat 
if we trace along the coast the level of the marine intercalation, 
which is supposed to be Oligocene, we shall find that it rises and 
falls in a gently undulating curve from the mouth of the ParaniL to aonth 
Patagonia : this is in correspondence with the fact that one flat basin lies 
beneath the Pampas, another beneath south Patagonia, and that in some 
places the Oligocene beds dip beneath the sea. The undulations thiou^out 
this great distance are, however, so extraordinarily flat that we may perhaps 
regard them as representing merely an original inequality of the sea floor. 

The Tertiary beds of Patagonia repose on the Cretaceous. In this 
case also the marine deposits are more recent as they lie nearer to the 
coast; here also great and exceedingly uniform oscillations have taken 
place, giving rise to an alternation of terrestrial and marine deposits; 
and here, once more, the diminishing amplitude of the positive phases 
conducts us to the existing state of things, while at the same time an 
inequality of great interest occurs in the movements. 

Upon the Cretaceous formation, and not sharply defined from it, rests 
the unf ossilif erous Qwiranitic stage : it is overlaid by a thick group of 
red sandstone and sandy clay, with gypsum ; this occupies wide areas in 
the west, towards the slopes of the Cordilleras (Piso Pehuencke with 
Meaotherium Marehii). To this stage we must perhaps refer the lignites 
of Punta Arenas. Next, for the first time since the Cretaceous period, 
a positive phase occurred. The deposits of this marine stage are exposed 
at many localities on the coast and in the interior: they everywhere 
contain the same mollusca, over their whole extent, from Paran& to Ponta 
Arenas, where they rest on the lignites (Piso Paranense with Orfjw 
Ferraresi) : they are correlated with the upper Eocene of Europe. The 
strand again receded, and beds of sandstone were laid down, with renuuns 
of plants and fresh-water mollusca, as well as mammals, among them 
genera closely allied to Palaeotherium and Anoplotherium of the gypsum 
of Montmartre (Piao Meaaptamico with MegaTnys PatagonieThsiay 

^ Informe oficial de la Gomision cientifica agregada al expedicion al Rio Negro bigo 
Tdrden de General Don J. A. Boca, 4to, Buenos Aires, 188S, III, Geologia, por el 
D. Doering, pp. 401-530. 



t :' 



*m 



', ' I 



i.' 



1 






CH,vn] THE TERTIARY LAND OF PATAGONIA 307 

Then for the second time the strand rose, but the marine deposits did 
not extend so far mland as in the preceding transgression : they probably 
correspond with the Oligocene (Piso Paiagonico with Ostrea Patagorda), 
The most westerly point at which the first marine stage occurs is Lago 
San Martin, near the CJordillera (lat. 49**-50^ S., west of long. 72' W.), while 
this second marine stage only attained, generally speaking, a half to one- 
third of the distance which intervenes between the sea and the Cordillera. 

Then the strand retired for a long distance and for a very long period ; 
several stages of continental deposits follow one another, uninterrupted 
by marine intercalations. Great quantities of trachytic detritus, which 
in the south occur in the upper part of the second marine stage, are 
continued into the lower part of the succeeding continental deposits. 
These begin with sandstone and marls : they contain the genus Anchitherium, 
indigenous in Europe during the lower Miocene (Piao Araueano). Then 
comes the alluvial sand of the western Pampas {Piso Pudche) : at the dose of 
ikU stage Doering places the limit between the Miocene a/nd PUocene, which 
coincides with the grecUest recession of the strand. The continent then 
extended much further to the east than at present. We have now reached I 

the Pampas formation, in the strict sense of the term ; it consists of clay 
with numerous remains of terrestrial mammals, and is subdivided into 
three stages — Piso Pa/mpeano inferior with Typotheriv/m, Protopithecus ; 
Piso EolUico ; and Piso Pa/mpeano lacustre. It is supposed to represent 
the Pliocene. 

A vast sheet of pebbles and conglomerate was now spread over the 
Patagonian plateau: it probably came from the Cordillera, and stands in 
some relation to the Glacial epoch. Erosion of the valleys then took 
place, and not till this period did the sea appear for the third time {Piso 
queraTuiino). It did not, however, extend far beyond the coast, least of 
all in the north, and the fauna corresponds very closely with that of the 
present day. This recent post-glacial transgression is thus separated from 
the two earlier transgressions, which contain none but extinct species of 
molluaca, by a great hiatus^ an interval wholly imrepresented by Miocene 
or Pliocene deposits. 

Doering adds that to discover the missing members between the second 
and third transgressions, we should have to make borings off cape Cor- 
rientes beneath the sea. 

In spite therefore of the great uncertainty as to the precise age of the 
beds, there is a remarkable concordance between the results obtained in 
three regions widely remote from each other by three able investigators, 
working independently. Neumayr of Vienna finds that at the beginning 
of the Pontic stage, which is generally recognized as the limit between 
the Miocene and Pliocene, the strand in the Mediterranean occupied a 
lower level than at present, and that in the Pontic period, or at the 

X 9 



308 TERTIAEY SEAS [part in 

beginning of the Pliocene, there is a gap in the series of marine bedLa. 
Heilprin of Philadelphia says that on the AUantie coast of the United 
States the Pliocene is not represented, and that the Miocene beds are 
separated by a gap from the succeeding and much younger shell beds. 
Doering of Cordoba arrives at the result that on the ooaat of Patagonia 
the strand must have retreated far from its present position, that the land 
must have attained its greatest extension at the limit between Miocene 
and Pliocene, and that the Tertiary, probably Oligocene, marine beds are 
separated from the post-glacial shell beds by a great gap. 

We will now return to the quer- Andinian shell beds. These are the recent 
deposits which were so clearly described by Darwin. Doering. who has 
traced the regular oscillations they record through 20 degrees of latitude, 
has no doubt that such movements could not have been caused by dis- 
placements of the solid land. A very remarkable p^pnliarity of the 
movement is its increase towards the aouth. In the bay of Plata the 
quer- Andinian shell beds occur at +20 to 80 meters; rising southwards 
they reach + 100 meters in the extreme south, where they are surn>undcd 
by terraces at + 300 to 400 metres. We shall revert to this &ct later on. 

6. Becent Li/mestane FarmatioThs. Formidable difficulties are con- 
nected with the question of the present deposition of limestone, and 
particularly with the mode of formation of those coral islands, so fre- 
quently ring-shaped, which rise from the great Ocean depths. 

DaLiTdisti^edth^^ 
reefs, which adhere closely to the land or a reef of rocks ; next^ barrier 
reefs, separated from the land by a deep chazmel ; and finally ring-shaped 
atolls, surrounding a lagoon, within which no land is visible. He attempted 
to explain the transition from group to group by assuming a gradual 
subsidence of the sea floor. Dana also adopted this view, as the result of 
very extensive observations of his own. Whatever opinion we may form 
as to the subsidence theory of Darwin, it needs no proof to show that 
a rising of the sea in the torrid zone would meet the facts in preciself 
the same way as a subsidence of the sea floor. Both Daiwin and Dana 
were well acquainted with a great number of apparently elevated coral 
islands, and they sought to distinguish between zones or r^ons of elevotioD 
and subsidence ^. 

Wilkes, the leader of the American expedition, which Dana accom- 
panied, expressed himself against Darwin's explanation, influenced appa- 
rently by having seen a number of elevated reefs, i.e. reefs laid dry by 
negative movement; and in 1865 J. C. Boss, who had lived for many 
years on Cocos island (Keeling atoll), the starting-point of Darwin's 

1 G. Darwin, The Structure and Bistribution of Coral Reefs, 2nd ed., 8vo, 1874 ; J. D. 
Dana, Corals and Coral Islands, Am. Joum. sci., 1885, 8rd ser., XXX, pp. 89-105 and 
169-191. 



CH. VIX] 



FOEMATIONOF CORAL EEEES 



309 



observations, opposed bis theory even more strongly. In the first place, 
Ross maintained that Darwin's statements regarding recent subsidence on 
Goods island were incorrect; and subsequently H. Forbes also asserted 
that in this case Darwin had been mistaken. The paper published 
by Ross on this subject is marred by much obscurity of thought as 
regards volcanic activity, and by other defects, which may be lightly 
passed over as the result of a lengthy ^residence on remote islands. On 
the other hand, it contains some valuable observations ; Ross remarked 
that all the elevated coral islands of the Pacific rise to about the same 
level, and that it appears as if the whole surface of the Ocean had 
suddenly sunk in consequence of the subsidence of one or several parts of 
the sea floor towards the centre of the earth. He further points out that 
Darwin represents elevation and subsidence as distributed in long separate 
zones, whereas elevated islands (reefs laid dry) and, according to Darwin's 
views, sunken islands (atolls) occur on the same line, close beside each 
other, and in between each other. He concludes that the distribution 
of coral reefs is determined by marine currents, which bring them 
nourishment K 

Finally, Semper, Rein, and Murray have of late years questioned the 
validity of Darwin's subsidence theory, because, in the opinion of these 
observers, the growth of the corals is always confined to the outside of 
the reef, where food is to be found ; while by the death of the interior 
part ring-shaped forms are produced, for which no subsidence of the ground 
is required 2. 

This question, which has such an important bearing on the statics of 
the sea, is discussed now because in various parts of the world limestones 
of recent formation occur in association with others, which are of late 
Tertiary age, and it is sometimes difiicult to discover a strict limit between 
them. Examples of this association are far from rare: on Antigua the 
beds which reach the sea are Miocene, and it ought not to be difficult to dis- 
tinguish them from the younger formations (II, p. 136, Fig. 15). The problem 
is more complicated in the case of the limestone of Chira^oa, Aruba, avd 



^ J. C. Ross, Review of the Theory of Coral Formations set forth by C. Darwin, &c., 
Nat. Tijdsclir. Ned. Ind. Yer. Batavia, 1855, YllI, pp. 1-43 ; H. Forbes, Notes on Keeling 
Idand, Proc. Geogr. Soc, 1879, pp. 777 et seq. 

* C. Semper, Die Riffe and das Leben im Meere, Zeitschr. wiss. Zool., 1863, XIII, 
pp. 563-569y printed in his Die Plulippinen und ihre Bewohner, 8yo, Warzburg, 1869, 
pp. 100-109 ; by the same, Die natHrlichen Existenzbedingangen der Thiere, 8yo, Leipzig, 
1880, II, pp. 89-98 and 261 ; J. J. Rein, Beitrftge zur pbysischen Geographie der 
Bermnda-Inseln, Ber. Senckenb. naturf. Ges. Frankf. am Main, 1869-1870, pp. 140-158 ; 
by the same. Die Bermndas-Inseln und ihre Korallenriffe, nebst einem Nachtrage gegen 
die Darwin'sche Senkongstheorie, Verb. I. deutdch. Greogr.-Tages zu Berlin, 8vo, 1882, 
pp. 89-46-; J. Mnrray, On the Stmctare and Origin of Coral Reefs and Islands, Proc. 
Roy. Soc. Edinb., 1879-1880, X, pp. 505-518. 



310 TERTIARY SEAS [pabtoi 

Venesfuda, which has been investigated by Lori^ \ In Florida the com- 
plication is still greater. Species pass into new forms or become extinct; 
those corresponding with still existing species increase in number, and 
the marine fauna of the West Indies is itself, like that of the existing 
Mediterranean, an assemblage of elements of different age and origin. 

For the purpose of a general survey it will be convenient first to refer 
to some observations in the Atlantic region, — Florida, the West Indies, and 
the Bermudas in particular, and next to discuss the chief structural featnres 
of the widely distributed coral reefs of the Padfia 

The peninsula of Florida in the north and north-west is formed, as we 
have several times observed, of the middle Tertiary Orbitoides limestone, 
which towards the Ocean rests against a zone of Miocene coming from the 
north ; to the south younger beds of limestone appear, till in the Everglades 
the surface of the plateau stands nearly level with the existing surface of 
the sea. Still further to the south the plateau forms the ragged shore line 
of the bay of Florida ; the shallow waters of the bay, only a few fathoms 
deep, cover a broad horizontal surface of calcareous mud, a region where 
material is at present accumulating for the formation of a new limestone 
plateau, similar to that of the existing peninsula. 

The bay of Florida, and Key Biscayne bay, its continuation to the 
north, are boimded on the Ocean side by a long series of low islands, the 
Keys, which run in a regular curve from the southern point of Virginia 
Key in the north, first to the south, then in an arc to the west up to the 
bank outside the Marquesas, through which it proceeds and reaches tiie 
Tortugaa As early as 1868 E. B. Hunt showed that this curve is pro- 
duced by the counter current coming from the north, which flows between 
the land and the Gulf stream. The counter current limits the zone of coral 
growth, since it drifts sediment from the north, and brings food to the 
corals. Alexander Agassiz has given a highly instructive account of the 
facts, and has shown how storm and waves attack the organic stmctores, 
break off fragipents, and grind them to a white impalpable powder, which is 
driven during the storms along the Keys and through the gaps into the 
broad quiet waters of Florida bay, where it is spread out by the play of 
the tides and accumulates on the sea floor as a fine calcareous silt. On the 
western side of Florida this accumulation of calcareous sediment also oocnis 
and extends far to the north, covered above by a thin crust of corals. These 
are frequently killed by the stifling mud which finds its way into their 
calyces. It is not, Agassiz remarks, that the sea floor has risen to a height 
at which corals could settle on it, but that the sediments have accumulated 
up to this height. As to the Keys, the Tortugas, and above all the 
Marquesas, a more or less annular form, which marks them as true atoUs, 

^ J. Lori^ Foasile Mollmken von Cnra^ao, Aruba nnd der Ettote von Venemela; 
SamoiL geol. Reichsmns. Leiden, 1887, 2. Ser., I, pp. 11K149, pi. 



CH. vn] SOMBRERO 

lua been {nrodoced solely by the inflaence of the carrente, wl 
nutriment to the coral polyps ^ 

Here then we h^ve before as at once onr first example of ; 
pUteaa now in process of formation, and an illust^iition of tb( 
which have been raised to the theory of the origin of atoUa by 
of the sea floor. 

Similar accomalations of calcareona mad and dead organic 
known over large areas in other parte of the West Indian r^o: 
for example, and particnlarly along Uie Tertiary plateau o: 
Alaeran r«ef (Scorpion rock) is a horseshoe-shaped atoll crowi 
calcaieons sediments ; it has doubtless arisen, like the Tortugai 
qneeas, without any appreciable displacement of the strand. 

In the West Indies signs of n^ative movement may be see 
I^aces. I do not allude to the limestone beds occurring at som 
tble altitude, the highest of which are found in Cuba and Ji 
ftltiwogh they have been regarded as reefs of comparatively r 
their age ia wholly unknown, and Uiey may even belong to the 
penod ' ; independent of these there are many indications of 
evrai close to the existing strand. Over the whole of the ea 
Florida, nearly aa far as the cranmencement of the Keys, hea a she 
the ' Coquina of St. Augustine,' which engirdles the coast and i 
as marking a n^ative mov^nent of 10 to 20 feet. This depo 
in many islands of the Antilles : a well-ki^own example is the 1 
of Guadeloape, which is surrounded by a recent coral reef. Y 
not be overlooked tbat the height above the sea of this last mi 
level surface does not afford a tjnistworthy measure of the : 
which have taken place during the existing phase (II, p. 26). 

Among the regions now laid dry in consequence of these mo\ 
mnat include the island of Sombrero, lying out to sea in lat. 
between the Virgin isles and the lesser AntiUes. 

Sombrero, according to Sawkins, is a little more tban a nanl^ 
length,iaid its greatest breadth is scarcely a third of its length, 
coast is bounded by a perpendicular or overhanging cliff, 25 to 4( 
with 12 to 14 fathoms of water outside. The greatest differen 
in the interior amounts to 10 feet : the surface is scored with 
fissures of erosion. 'A more desolate or inhospitable spot can 
foond in the actual world or in the realms of the imagination.' 

' A. AgaBoi, The Tortugas and Florida Reeb ; Mem. Am. Acad. 8ci., 
Volume, Cambridge, 1885, XT, pp. 107-188, mapi. 

' W. A. Croiby, On the elevated Eeefc of Cuba ; Proc. Borton Soc. Nat 
1S83, XXll, pp. 124-130. The tenocea of limertone lie at heighU of 30 i 
feet, and 500 feet ; even the limectons of the Ynnqne at a height of 1 ,800 t 
entirely different, w conridered hy Croab; to form part of the eame groap. 



TERTIAEY SEAS 



[PABT m 



The whole isla&d is formed 
UadreporM of limestone ; the mollusca en- 

closed in the limeatone belim^ 
to the existing fauna, with ihe 
exception of a very common 
Porites Bulla, which appears to be Bulla . 

Hadreporea gruTwea of the middle Tertiary 

Shells "' *'^* ^^ Indies. Dykes of 

phosphates occur in the lime- 
Btone, evidently derived from 
vaoifihed beds of guano. They 
CfTsUiiine j^ye been a sonroe of attractioi 

limeaUiue 

to this barren island since 1S56; 
a description of them has been 
given by Alexis A, Julien ', 
Six beds of limestone aie to 

8h lU d ^ dietingaished, resting hori- 

Hadreporea zontally one on the other; we 
will indicate them from the sea- 
level upwards by the letten A 
ioF. On the face of the cUffi, 
htjwever, only two bedding- 
planes, as a mle, are at all con- 
spicuous : they ore distinguished 
by partings of a different natnre 
which occur between A and B 
and between D and E, A further 
distinction is afforded by the 
colour ; D is decidedly grey, ud 

Shell* *^"^ stands out in marked con- 

trast to the other beds, which 
are pure white. Beds of guano 
have been formed, according to 
Julien, at three different epochs 

„ , at least : each of these beds, how- 

Hadreporea , , , 

ever, was subsequently removed 

by the sea, leaving some remuns 

in the clefts of the limestone. 

* Sawkine, Report on Jam&ic*, I, 
p. 261 ; Clere, Geolog; of the Wot 
Indian Isles, p. 21 ; A. A. Jnlien, On 

(So».feTel) thg Geologj of the Kej of Sombrero, 

Ann. Lye. Nat. Hiit N. Toik, 1B67, 

(after A. Julien). VIII, pp. 251-278, pi. 



White limestone 



Sheila, ht. 
White limestone 



BERMUDAS 313 

loneludeB from this, aa well as from the character of the limestone 
at the island had heen repeatedly submerged and again laid dry. 
attempted ib represent the oscillations graphically by a carve, and 
r the trifling number of beds here visible, was obliged to assume 
iTations and nine sabBidencee of the ground, 
resemblfuice of these beds to those of the eastern Alps is evident, 
r each single bed does indeed correepond to an oscillation or not, 
D be no donbt that we have before ns here formations which bear a 
likeness to the stratified limestones of earlier epochs, 
limestone beds which rise within the encircling coral-reefs in the 
idles are tnnnelled by many Ifirge caves. In the Bahamas, where 
of ancient strand-lines may be recognized, such caves occur in many 
Sharpies has described those of the Caicos idan<h in the neighbour- 
the Turk islands : the floor is covered with red earth, gypsum, and 
It seems to me scarcely possible to doubt that these terraced lime- 
izcavated by caverns, are older than the living reef around them. 
:eat age is confirmed by the fact that Pomei and Cope have found 
tins of extinct species of terrestrial animals in similar caves in the 
and of AnguiUa, and from this the great age of the negative move- 
vy be inferred. 

conclude ihea that small annular atolls without deep lagoons, such 
Tortugas trnd Alacran, may be formed without any appreciable 
movement of the strand. The existing Keys of Florida must have 
ilt up since the last negative movement. If a positive movement is 
progress it is so trifling that the sediment is able to keep pace with 
I negative movements and the caves with remains of extinct ter- 
mammals date from a more remote period. It is not improbable 
I separaiioQ of the limeatone into beds is due to minor oscillations ; 
TO seems to suggest this, but in the bay of Florida a change in the 
af the sediment might easily be brought about merely by the closing 
gement of the openings between the Keys, by increased evaporation, 
her ways, without oscillation of the strand. 

he Bffmvudae we are afibrded the rare opportimity of observing the 
ine structure of a group of calcareous islands. They have been 
id by Nelson, Jones, Bein, the members of the Challenger expedition, 
le '. A recent outer reef clings close to the south-east side of the 

, Sharpies, Torks leland and the Gnano Caves of the Cucob IilandR; Proc. 
k»c Nkt. HiBt, 1884, XXII, pp. 242-252. 

Kelson, On the Qeology of the Bermndas, Tnni. Oeol. Soc, 2nd «er., V, p. 108 ; 
in, Beiti&ge zur phyaischen Oeogiaphie der Bermnds-Inseln, Ber. Senckenb. 
lea. F»nkf. &m Main, 1869-1870, pp. 140-158; J. M. Jones, Recent ObseTvationi 
trmndM, Nature, Aag. 1, 1872, p. 262 ; W. Thomson, The Atlantio, 8vo, London, 
pp. 290-867; W. N. Rice, The Oeolc^y of Bermada, in Jones and Goode, 



314 TERTIAEY SEAS [pake in 

higher islands, but recedes from them on the south-west, as well as on the 
north, for a considerable distance, to form a broad oval ring, which thus 
encloses all the higher islands and an extensive lagoon. ' The islands attain 
a height of 250 feet, and are formed above the sea-level of drift rock, that 
is, of organic d^ris which has been piled up to this height by storms, and 
then either loosely cemented together, or converted into dense and compact 
limestone, by the rain. . Bed earth covers the islands, penetrates into the 
fissures of the white limestone, and forms the soil for vegetable growth. 
There are caves in the limestone with stalactites hanging from the root 
In some places indications of oscillations are to be seen ; at Stocks point 
Bice describes a bed of beach conglomerate, formed of masses of drift; 
rock, with large sea shells and compact lumps of red earth. This con- 
glomerate occurs as an intercalation in the drift rock, which contains 
land shells. 

At the mouth of Hamilton harbour submarine mining was commenced 
in 1870, and, according to Jones, a cave with stalactites and red earth "wbb 
encountered at a depth of 6 fathoms. Excavations were continued later on 
a larger scale, with a view to the construction of docks. At a depth of 
— 25 feet, a calcareous mud was met with, 5 feet thick; beneath this loosely 
cemented coral sand, with fragments of Diploporia and shells ; then followed, 
at a depth of —45 feet, a layer of peat moss or rather old vegetable soil, 
with upright stumps of trees, land shells, and bones of birds. This lay on 
ancient limestone. 

We have here definite proof of positive movement: nevertheless Bioe 
concludes, with a high degree of probability, that no appreciable change has 
occurred since 1609. If the movement is still in progress it must be very slow. 
Bice bases this result on a number of ancient observations collected by him* 

In passing to the Pacific region we may mention first that here, at 
least in the regions which border it on the west, there is no lack of evidence 
in proof of continuity between the very recent limestone formations and 
those of an earlier date. 

Tertiary deposits play a large part in the formation of the archipelago 
west of New Guinea, and recent limestone inunediately follows patches of 
Tertiary. The Aarv, archipelago, according to Biedel, includes twelve 
large and eighty-three smaller islands ; all form part of a single limestone 
plateau. Five very narrow channels of the sea, not broader than fairly 
large rivers, separate the six principal islands, which form together a 
plateau 180 kilometers long, with a slightly undulating surface, covered 
with marshes, in which brackish-water mollusca live. The south-east part 
of the island, which is the highest, and rises to 50 meters, is formed of late 
Tertiary beds (II, p. 166). 

Contribution to the Natural Histoiy of the Bermndas, I (Bull. U.S. Nat. Mub., No. 25), 
8?o, Washington, 1884, pp. I>d2. 



CH.vn] RECENT LIMESTONE PLATEAUX 315 

The BuTida pUxieau of South Australia shows to what dimensions an 
exposed platform of Tertiary limestone may attain. 

Tertiary fossils have been recorded even from Viti Levu, Fiji (II, p. 164). 

In the region of the Pacific coral reefs, as in the West Indies, there are 
numerous traces of negative movement. Tabular limestone masses with 
a horizontal surface rise out of the rings of living reef; sometimes they are 
terraced in steps, sometimes ruled along their steep cliffs with horizontal 
strand-linea Caves with stalactites run through them, sometimes for miles ; 
and red earth occurs on their upper surface. Masses of limestone such as 
these are met with here and there as far to the east as Henderson (Eliza- 
beth) island, beyond the Paumotu islands ; this, according to Beechey, is 
80 feet high, with steep cliffs composed of limestone which breaks with a 
oonchoidal fracture like a Mesozoic limestone ^ Nowhere, however, does 
the height of the plateau appear to exceed 100 meters. We do not know 
what part blown sand or drift rock plays in the structure. 

In the south part of the Pdew islands the negative signs become so 
marked that Semper was inclined to attribute the origin of the group, not 
to subsidence as Darwin did, but to an elevation of the sea floor. The 
limestone clifis here attain a height of 250 feet, and end above in a perfectly 
horizontal line ; towards the east they are 80 feet high, and again horizontal 
at the sununit K 

On the slopes of the islands in Bougainville straits, in the Solomon 
group, regular terraces occur, which Guppy regards as so many elevated 
barrier reefs ». 

One of the most striking examples is afforded by the Loyalty group, 
which borders New Caledonia, like a second line of coast. The islands 
form a long series. The first is Astrolahe, which is just awash. The 
second is Uvea, a circular island formed of coral limestone, rising in a single 
step to a height of 15-18 meters. In the centre lies a lagoon with a very level 
floor and a maximum depth of 18 meters ; the island is quite horizontal, 
and terminates against the sea in an overhanging cornice. 

The third island is Llfu ; this has no interior lagoon ; it is a plateau, 
rising in three horizontal stages to a height of 90 meters. For thirty 
nautical miles, from one end to the other, each stage maintains precisely 
the same leveL Great caves are to be seen. The sides of the plateau 
sink to unfathomed depths, as is the case with all these islands. So 
Chambeyron described it; Clarke, who gave an account of lifu in 1847, 

^ Captain F. W. Beechey, Narration of a Voyage to the Pacific and Beeiings Stiaita, 
4to, London, 1831, 1, pp. 55-58 and 187. 

* Semper, Ezistenzbedingangen, 11, p. 76. 

' H. B. Gnppy, Suggestions as to the Mode of Formation of Barrier Reeft in Bongain- 
TiUc Straits, Solomon Group ; Proc. Linn. Soc. N.S.W., 1884, IX, Sydney, 1885, pp. 949- 
959, pi. 



316 TEBTIABT SEAS [past ni 

found evidence of two elevations, one of 80, the other of 170 feet, or 
together of 250 feet. Balausa mentions the red earth, without which 
vegetation could not exist, and recognized four atoges, of which the third 
runs right round the island. There are wells in the limestone, 41 meteis 
deep ; it shows no stratification. 

The fourth island is MaH ; it presents, according to Chambeyron, 6tc 
stages, all horizontal and still more clearly defined than in Lifa ; the upper- 



Fio. 30. Uvtat, LofdUti Ormv (alter Chunbeyron). 

most plateau reaches a height of 90 to 100 meters. Between the second and 
third stage lies a broad plain. Id the north-west of the island the plateaa 
of the highest stage occurs ; here a peak of volcanic rock protrudes. 

Towards the north-west the Loyalty group ends in the Petrie reef, and 
to the south-east in the Durand reef and Walpole island, the last an isolated 
rock 95 meters high. 



Flo. 81. Lifii, LoyaBy fjroup (after Chamb«7roD). 

It may be asserted that each of these islands, while maintaining its 
honzontality, has been raised spasmodically to these heights; Har^ to 
90 to 100 meters, IJfu, 90 meters. Uvea, IS to 18 meters, and Astrolabe 
not at all. For my own part I cannot regard this as probable. I should 
prefer to suppose that to begin with thete was a oontinuous limestone 
platean, + 90 to 1(X) meters in height, and at this time the sea-level stood at 
least 90 to 100 meters higher than at present. Of the highest part of this 
plateau one fragment has been preserved in Mar^ close to the volcanic peak, 
and another in Lifu; the lower stages mark a cessation of the negative 
movement or a phase of the positive recurrence, or perhaps merely the 
intercalation of a less resistant bed. The overhanging cornice of TJvw 



cH.vn] METIA 317 

shows how the breakers are engaged in their work of destruction to-day, 
as they have always been in the past \ 

Finally, mention may be made of the island of Metia, which lies north 
of Tahiti. According to Dana it is 250 feet high, and consists of coral 
limestone with great stalactitic caves ; on its summit is a plain, similar to 
the beach platform of existing coral islands ; isolated sea stacks show that 
a part of the limestone has been removed by the sea. Two horizontal lines 
ran round the brow of the cliff ; along these caves are frequent ; the rock in 
ooDsequence presents a stratified appearance, and is divided into three 
ahnost equal parts. It appears to me a fact of great interest that of the 
two specimens gf the rock which Dana took from Metia, one contained 
38^07 and the other only 6-29 per cent, of 'magnesium carbonate. Here 
dolomite and limestone have been deposited in the open sea, as in the Flatten- 
kalk of the Alps. Dana regards Metia as the core of the deposits which 
filled up a once existing lagoon, and thinks that the carbonate of magnesia 
may have been precipitated in this lagoon as a consequence of evaporation '. 

Far more numerous than these comparatively elevated islands are those 
of inconsiderable height, particularly such as are bordered by an overhang- 
ing cornice, which rises no more than a few meters above the sea-level. 

Attempts have been made to explain the inequality in height of these 
islands by an unequal elevation or subsidence of the various regions in 
wfaidi they occur. This is a view I cannot share. The existence of the 
limit of 100 meters, which is rather strictly maintained, is opposed to it, 
as well as the occurrence of repeated terraces or strand-lines. 

These facts tend rather to show that all these limestone plateaux are 
the remains of older formations, certainly older than the living ree& of the 
present day; like the similar limestone plateaux of the Bahamas, which 
are also excavated by caves and covered by red earth ; and probably all 
the negative movements to which they bear witness have affected uniformly 
the entire region of these coral islands, and belong to a more remote period. 
We do not know the age of these higher limestone plateaux. Dana has 
often expressed the opinion that the epoch of the Pacific coral reefs 
embraces the whole Quaternary, and perhaps a part of the Tertiary period. 
Consequently it is not to these plateaux that we should turn .for an 
explanation of the existing state of things ; this can only be obtained by 
a study of the living ree&. 

^ Gaptain Ghambeyron, Note relative k, la Nouvelle-Caledonie, Bull. Soc. g^ogr., 1875, 
^ 8^., IX, pp. 566-586 ; W. B. Clarke, On the Geology of the Isle of LafQ, Quart. Joum. 
^1. Soc., 1847, in, pp. 61>64 ; B. Balansa, Nouvelle-Gal^donie : les flea Loyalty, Bull. 
^ g^gr., 1873, 6« 8^r., V, pp. 521-584 ; Grundemaim, Peterm. Mitth., 1870, XVI, 
pp. de5>869. Ghambeyron gives Lifu a height of 90 meters in the text, of 60 meters 
in the diagram ; the first figure corresponds with that of other observers. 

' Dana, Coral Islands, pp. 198, 857; Bibourt, Observations g^logiques sur Tahiti et 
les ttes basses de Parchipel des Fbumotns, Bull. Soc. g^ogr., 1878, 6« s^r., XVI, p. 85. 



318 TERTIARY SEAS [pabthi 

Darwin was well acquainted with the more active growth on the outer 
side of the reefs, and Dana even remarks that a coral reef is a limestone 
plateau with a living edge. Nevertheless it appears from recent observa- 
tions that the influence of the currents, which bring nourishment, is much 
greater than Darwin and Dana supposed. They may of themselves give 
rise to annular reefs of no great size ; this is shown by the little atolls of 
Florida, and we must probably refer to this cause the small atolls of MaJdos 
Madhu in the Maldives, which are grouped together into larger rings; 
Darwin has fully described them. The little goblet-shaped reefis of Serpulite 
limestone in the Bermudas belong to this clasa But Alexander Agassis 
himself has expressly admitted the difficulty of explaining lagoons of great 
depth, without conceding a change of level ^. Semper^ in a section through 
BdbeUhaiiby the largest of the Pelew islands, has estimated the depth of the 
lagoon at 50-60 fathoms, and numerous soundings in lagoons have reached 
these figures. According to all observations so far made reef-building 
corals cannot live at this depth. 

This difficulty then stands opposed to all the theories advanced since 
the time of Darwin and Dana, and still retains all its force as one of the 
weightiest arguments in favour of the so-called theory of subsidence. Two 
objections might be raised to this theory : first, that it is not possible to 
understand how so large a portion of the earth's surface could have sunk 
so slowly and uniformly ; and next, as Boss pointed out, that among the 
islands asserted to have subsided stand fragmentary plateaux, which have 
apparently been elevated. But these facts will no longer appear strange, 
as soon as we admit that it is not the land but the sea which is the variable 
element. 

As regards the actual amount of the positive excess of the oscillating 
movement, no estimate, not even approximate, seems to me as yet to be 
possible. The depth of the lagoons implies positive movement, but there 
is a wide difference between the 40-60 fathoms of a lagoon (which may, it 
is true, have been filled up to an unknown extent), and the very consider- 
able depths which have been sounded, not indeed everywhere, but yet in 
many places, in the immediate neighboiu*hood of atolls, and which reveal 
very steep slopes. 

That the atolls stand on a rocky base has never been denied. Darwin's 
description of the relations of barrier reefs to atolls involves this, and the 
many visible peaks of volcanic mountains testify to it. 

Murray has shown what vast quantities of pelagic organisms with 
calcareous shells occur in certain parts of the Ocean ; the shells after the 
death of the animal fall like rain to the bottom, where they are dissolved 
by the carbonic acid of the sea- water at great depths, but accumulate in 

^ A AgaBsiz, The Toitugas and Florida Beefs; Mem. Am. Acad. ScL, Centennial 
Volume, Cambridge, 1885, XI, p. 121, maps. 



^ 



vn] GUANO ISLANDS 319 

moderate and shallow depths. I believe, however, that the eases are very 
rare in which this kind of sediment plays an important part in the stracture 
of the substratimi of atolls. The mantling over of submarine mountains 
with limestone is ill suited to explain the isolated atolls in the middle of 
the Ocean, and volcanic cones would by this means very seldom reach the 
seme of coral growth. Murray's casual reference to Oraha/m idarid (Julia 
or Ferdinandea island) seems to me much more fortunate. 

This island made its appearance in 1831 as an ash-cone on the south- 
west of Sicily, in a sea 100 fathoms deep. The waves attacked the mass 
of loose material, and by the end of a few months a large part had been 
washed away ; the central pipe, filled with scoriae and ash, was exposed. 
The sea continued its work, and after a short time nothing was left but 
a slightly submerged bank, which, secure from the destructive action of 
the breakers, persisted for a long time. In this way a foundation may be 
prepared at a suitable depth for the settlement of corals. Then all those 
processes would be set in train which have been described by Semper, Bein, 
and Murray, and most important the more active growth of those indi- 
viduals which occupy the margin of the ring. 

Some such explanation might possibly hold for the little guano islands, 

lying near the middle of the Pacific. It is true that some of them, such as 

Jervis, M'Kean, and Hero islands, are supposed to show signs of negative 

' movement, but the only evidence for this is the presence of gypsum in the 

lagoon beneath the guano. 

Jervia island (lat 0"* 22^ S. ; long. ISQ"* 58' W.) was examined by Hague. 
A ring-shaped rampart surrounded by a fringing reef rises to a height of 
-f 18 to 28 feet. The plain within the ring lies at a height of + 7 to 8 feet ; 
it consists of a layer of gypsum covered with guano. In its deepest part 
gypsum and sea salt occur^. 

On the other hand, Dixon writes of Maiden island (lat. 4"* 2' S. ; long. 
154^ 58' W.) that the sea-water percolates through the reef itself into the 
lagoon, and there evaporates; an outfiow only occurs at very low water. 
Bare but violent showers of rain wash the salt out of the lagoon but leave 
the gypsum behind. The rampart attains a height of + 21 feet. Ancient 
kitchen middens are found on this island. Several concentric ridges, formed 
of loose blocks of reef rock, succeed each other both on the north and south 
side of the island ; they run parallel with the shore, and have been piled up 
by waves of imusual power *. 

Not only the waves produced by storms, but those due to earthquakes, 
which travel from time to time across the Ocean, must doubtless leave 

* J. D. Hague, On the Guano Islands of the Pacific Ocean ; Am. Joum. Sci. Arts, 1862, 
2nd ser., XXXIV, pp. 224-243, in particular p. 230. 

* W. A. IHxon, Notes on the Meteorology and Natural History of a Guano Island ; 
Jonm. Roy. Soc. N.S.W., 1878, XI, pp. 166-175. 



320 TERTIARY SEAS . [PAirrm 

their traces on the strand. The occurrence of negative positive movement 
in these isolated spots cannot therefore be regarded as ascertained. 

It follows from the foregoing observations that the substnictore of the 
coral islands is of various nature and age. Yolcanos are still active, and 
new bases for the attachment of corals are continually being formed. Bat 
the settlement of the polyps on the summit of a few isolated ash-cones 
can only account for some of the reefs. Many may be seen resting on 
fragments of ancient continental land ; from others the peaks of volcanic 
masses project, showing that loose ash-cones have not formed the only 
basis of support. The deeply eroded valleys of these peaks may in all 
cases, according to Dana, be regarded as an indication of positive move* 
ment. The rule that in the Atlantic Ocean the islands are arranged in 
straight lines, and in the Pacific in arcs, has its cause deep in the original 
structure of the globe (II, p. 205). It applies as much to fragments of 
folded Cordilleras as to the alinement of volcanic series and the distribution 
of coral reefs. The volcanic line of Fernando Po is straight, and the lines 
of coral reefs in the Maldives and Laocadives are straight also. In the 
west Pacific all the islands, whatever their constitution may be, are 
arranged in arcs. Thus the coral reefs disclose to us the plan on which 
the heights beneath the sea are distributed, or, in other words, hy means cf 
the coral reefs an isohypda of the awbmerged mountains of the Ocean is 
projected on its surface. This important fact affords, however, strong 
support to the theory that the coral reefs have been produced under the 
influence of positive movements, and there is nothing to indicate that the 
peaks and crests of the submarine mountains are of equal height. However 
frail may be the arguments on which Darwin and Dana have based tiieir 
rough estimates of the thickness of such structures, yet it must be observed 
that the figures they obtained, viz. 1,150, 1,750, and 2,000 feet, are well 
within the thickness of the stratified masses of limestone and dolomite, 
which accumulated over a large region of the Alps during the epoch of the 
Plattenkalk alone, i. e. of an upper division of the Eeuper. But that the 
sea of the Plattenkalk, at least towards its close, was subject to oscillations 
with a positive excess seems to me to have been establielied. Oscillations 
may be deduced from the precise observations made by Rink on the 
Nicobar islands, by Junghuhn and von Richthof en on the reef of Udjmig- 
Tji-Laut-urun on the south coast of Java, and by von Drasche in the 
neighbourhood of Paracali on the east coast of Luzon. In the last two 
cases a dead reef may be seen separated from the land by a plain of oonl 
sand. Outside the dead reef is the living reef bathed in the wash of the 
surf ^. I regard all the tabular masses of dead coral rock, which rise to 

* H. Rink, Die Nikobarischen Inseln, 8vo, Eopenhagen, 1847, p. 82 et seq. ; JanghaliOf 
Java, III, p. 1442 ; BIchthofen, Zeitschr. deatsch. geol. Ges., 1874, XXVI, p. 240 ; Drasche, 
Luzon, p. 62, and fig. 14. 



CH. vn] ANCIENT CORAL BEDS 321 

a height of 100 meters and extend as far as the remote Henderson island, 
as indications of ancient oscillations. Unfortunately, we do not possess 
any detailed description of the organic remains found in these islands. In 
the Atlantic, outside the region of the ree& and to the north of them, sedi- 
ments have been laid bare in the same manner on the Azores and Madeira ; 
and these have been referred to the first Mediterranean stage, i. e. to one 
of the lower subdivisions of the Miocene. In Madeira (lat. 32° 43' N.) they 
rise to a much higher level than the fragments of the Pacific plateaux, 
attaining a height of 1,350 feet (I, p. 288). 

Darvrin, I think, went too far when he attempted to distinguish r^ons 
of elevation from those of subsidence by plotting the distribution of 
fringing reefs, barrier reefs, and atoUs; Dana has already expressed the 
same opinion, but the attempt which Dana also made to determine regions 
of maximum subsidence according to the number and size of the atoUs 
does not rest on a surer basis. On the other hand, in spite of the extremely 
valuable information accumulated by recent observers, I believe, in agree- 
ment with F. von Bichthof en, that we must still accept as a true explana- 
tion the fundamental idea of the subsidence theory, according to which 
the larger coral reefs have been built up imder the influence of a widely 
distributed oscillation of the strand, with a positive excess K 

This movement, if it still continues, must take place so slowly as to be 
beyond the reach of measurement, so slowly indeed that little reefs, like 
the Tortugas, the atolls of Mahlos Mahdu and other islets, may come into 
existence without betraying any signs of its action. 

Thus we have reason to believe that theoe reefs have arisen under the 
influence of a predominant positive movement^ and on the other hand 
those who maintain that no movement is taking place aJt the present day 
are right, in so far as no such movement can be directly demonstrated. 

It has been repeatedly asserted that Darwin's views are not in accor- 
dance with what is known as to the mode of occurrence of corals in ancient 
marine deposits. I have made a personal study of all the more important 
and well-kiown coral formations in the Gosau beds, the Eocene of Cormons, 
the Oligocene of Crosara and Castel Qomberto, yet notwithstanding the 
abundance and variety of the corals, sometimes forming really large isolated 
growths, I have never seen anything in these localities which could be 
called a true coral reef. The corals lie heaped together, with shells, in 
tufis or marls, that is, always in clastic sediments. The loose nature of this 
material renders the corals an easy spoil, consequently they are abundantly 
represented in collections, and hence the reputation of these localities. 
Coral reefs must be looked for in hard limestone, but in these late Neozoic 
formations I know of no masses of limestone which could be termed reefs 
in the true sense of the word. The only conclusion to be drawn from this 

^ F. von Richthofen, FOhrer fOr ForschungsreiBende, p. 406 et seq. 



322 TEETIAEY SEAS [PAirrm 

is that at this period the requisite conditions for the growth of reefs did 
not exist in Europe. 

In the Rhaetic stage there are marly beds which have aflforded corals, 
and the Rhaetic zone of the southern Alps contains reef-boilding spedee, 
such as Convexaetrasa Azzarclae, ThaTriTiaetraea Meriani^ and Adraeo- 
fMyrphx Sastianiy which also occur in the northern Alps^. But these 
species, biu'ied in marl, do not form coral reefs. To discover anyihing 
like continuous growths we must examine the underlying or intercalated 
lithodendron limestone ; here we may see masses 15 to 20 meters tiiiek, or 
perhaps even more, which are actually formed of the cylindrical oorallites of 
Lithodendron ; but these also, as far as I have been able to observe them, 
have rather the appearance of thick beds than of true reefs. The arrange- 
ment in beds is predominant, and the resemblance to the beds of Sombrero 
or to the incrusting coral growth on the calcareous mud of Florida bay J& 
unmistakable. The evidence presented by these ancient sediments of the 
deposition of dolomite, as such, directly from the sea is, however, of 
the clearest nature. Agassiz says that the fine calcareous mud suffocateB 
and kills the corals. On the Osterhom corals occur in a white dolomite, 
and their calyces are filled with dark coloured dolomite, which has entered 
them from above *. 

The existing structures which are most closely allied to the coral beds 
of south Tyrol are perhaps the older tabular masses which now project 
from the middle of recent reefs, but there is no reason to suppose that the 
beds in south Tyrol were surrounded by similar reefis. With the exception 
of those places where interbedding with adjacent tuffi occurs, and where 
we appear to have to do simply with an interrupted formation of limestooe, 
these tabular fragments had the form of islands. 

6. Oevieral Survey. We have referred to only a very few of the remark- 
able events which distinguish the history of the Tertiary aera. Towards 
the close of the Cretaceous period, but before it had actually come to an 
end, a great and progressive decrease in the ext^it of the sea b^an to 
take place. In the middle of the central Mediterranean and in the Sahara 
we can discern no trace of this ; there, marine sediments succeed each other 
without interruptioiL But around these regions large areas were laid dry, 
fresh-water formations accumulated : on the one hand the Oarumnian seiieB, 
which extended on both sides of the Pyrenees and into the valley of the 
Rhone, on the other the contemporaneous libumian series, which covered 
the region north of the Adriatic; these show how great the negative 
movement must have been, and recall the conditions which prevailed at 
the dose of the Jurassic period. The Tertiary sea then advanced by 

^ Renss, Ueber einige Anthozoen der EOssener Scbichten und der alpinen Trias; 
Sitznngsber. k. Akad. Wiss. Wien, 1864, L, pp. 158-168. 
' Cf. note 8, p. 263 of the preceding chapter on the suffocation of the corals. 



J 



CH, vn] 



EEVIEW OF TERTIAEY SEAS 



823 



repeated oscillationB : in the Oligocene a temporary oonnexion between 
Europe and the far north was established along the east slopes of the 
Uial; then with ev^-increasing clearness we perceive the progress of 
those events by which the ancient central Mediterranean sea was dis- 
membered and diminished. Its oommunication with India was destroyed. 
Then it lost Irania, Turkestan, Asia Minor, and the western border of the 
Alps; the conunmiication, indicated by the sub-littoral fauna, with the 
West Indies was brdcen up. Then the Sannatian region became separated 
off; finally^ not only the Aralo-Caspian region, together with south Russia, 
but also the valley of the Danube was given up. Next came a period of 
erosicHi, when valleys were carved out, and then the Pontic Cardium beds 
were spread abroad as far as the valley of the Rhone. This is the epoch of 
greatest shrinkage : the strand stood lower than at present. Then the sea- 
level began to rise again, and at the arrival of the northern immigrants 
it exceeded its existing height. Subsidences extended the area of the sea, 
and the world became much as we see it now. The existing Mediterranean 
is the residuum of an Ocean, which extended parallel to the equator, and 
at one time, before the Atlantic came into existence, surrounded half the 
globe. 

There were other marine areas which maintained a protracted existence, 
but have left no recognizable signs of the manner in which they opened 
into the Ocean : these were situated over the regions bordering the North 
sea in the south, in particular the north of Germany and the south-east 
of England (I, p. 291) ; but in the whole region of the North Atlantic 
coast, as well as in North America down to the fortieth parallel, marine 
Tertiary deposits are wholly absent. North of the Lofoten islands, we 
enter a region which, as we found when studying the distribution of the 
MesoKoic seas, is more closely allied with the Pacific than the Atlantic: 
the Tertiary molluscan fauna of this region, which reveals the temporary 
advance of the sea across Spitzbergen to east Greenland, where it is 
associated with the plant-bearing beds, may perhaps have come from the 
Pacific Ocean by way. of Mulito on the lower Yukon. This is a point 
on whidi we do not possess as yet any precise information. 

The coast of North America, below the fortieth parallel, presents 
a regular succession of marine Tertiary beds, resting on the Cretaceous, 
which extends far towards the south; but here also there is a gap. 
Certain upper members of the Tertiary group are apparently absent, and 
on the other hand a very recent marine formation occurs in transgression. 
Here, also, we may fairly suppose that in a late phase of the Tertiary aera 
the strand-line lay lower than at present, and in the next succeeding phase 
it again stood higher than it does now. 

The Caribbean sea is also a remnant of that great Ocean which once 
extended parallel to the equator, across the existing Atlantic This is 

Y % 



324 TERTIAEY SEAS [pAirrm 

8hown by the presence of European elements which continued to be in- 
troduced into its successive marine faunas, up to the close of the first 
Mediterranean age. Thus it appears that those two parts of the Atiantic 
coast, which in contrast with the rest possess the Pacific stnicture, L e. tiie 
Cordillera of the Antilles and the broken arc of Gibraltar, mark the areas 
once occupied by that ancieni Ocean^ the ^Central Mediterranean^ of 
Nev/mayr. 

In the interior of the United States the negative phase, which closes the 
Cretaceous transgression, is represented by the Laramie stage ; great fresh- 
water lakes existed here during the different periods of the Tertiary aera. 

On the Patagonian border of the Atlantic extensive continental deposits 
with two marine intercalations extend to the sea ; they show that the shore 
was sometimes situated much nearer the cordiUeras, sometimes farther to 
the east than the existing coast. The second transgression did not extend 
so far as the first ; but a third marine formation follows the quer-Andinian 
stage, with a very recent fauna ; it rises with continually increasing height 
as it is traced towards the south. Between the quer-Andinian stage and the 
last marine Tertiary formation a great hiatus occurs, greater, so far as we 
can judge, than that of North America, and certainly much greater than 
that of Europe : and here, also, we find on the whole that at a late stage of 
the Tertiary aera the strand-line lay lower, and at a still later stage higher 
than at present. 

The Tertiary beds of Chili, of which Philippi has given instructive 
accounts, I have not ventured to take into consideration. The increase in 
the height of the quer-Andinian stage towards the south, which similarly 
occurs in Chili, will be specially considered later. Owing to the fragmen- 
tary state of our knowledge I have been compelled to refrain from a 
discussion of the Tertiary deposits of the Pacific region, and their continua- 
tion on the south border of Eurasia, in Java and Sumatra, along the Indiis> 
and in the Persian gulf. 

Within the Indian and Atlantic regions the Cretaceous covers the east 
of Brazil, and proceeds towards the coast without an accompanying Tertiary 
border. The sea washes against a strip of Cretaceous, Which extends, with 
interruptions, from Piauhy to Bahia, and probably to the Abrolhos. It is 
the same on the west coast of Africa, from the Elobi islands to Mossamides. 
In Natal, also, marine Cretaceous deposits advance to the sea unaccom- 
panied by a Tertiary border. The facts are the same in the north-west of 
India, on the Narbada ; and off the mouth of this river lies the little island 
of Perim, formed of Tertiary sand and gravels, with the remains of Dino- 
therium. Mastodon, and giraffes — an evidently fluviatile deposit. Near 
Pondicherry likewise the Cretaceous occurs without a Tertiary border. 
But associated with these intervals of coast there are others of still greater 
length, where neither Cretaceous nor Tertiary is seen, but ancient rocks 



CH. vn] NO PROOF OP CHANGE IN PROGRESS 326 

alone; this is the case in the south of Brazil, as far as and beyond Bio 
Grande do Sul, in Africa for an unknown distance, in India from the 
Narbada to cape Comorin, and from Pondicherry to the mouths of the 
Ganges. The Cuddalore sandstone in the south-west is of comparatively 
very recent date. 

According to all the evidence with which I am acquainted these coasts, 
distinguished by the absence of an outer Tertiary border, are found chiefly 
in low latitudes, forming a zone in the Atlantic and Indian regions. 

The coral islands of existing seas exhibit in many places relict strand- 
lines and relict fragments of ancient plateaux; they occupy a region 
characterized by very extensive and extremely uniform oscillations of the 
strand-line : the oceanic atolls have attained the surface of the sea owing 
to the existence of the positive excess. In Florida the foundation of the 
Keys has been formed by an accumulation of material under the influence 
of a current. The continuance of a positive phase at the present day 
cannot be demonstrated from the living reefs, nor is it susceptible of 
measurement. 

With the quer-Andinian stage we reach the most recent transgression 
known to us : the coral reefs are the result of processes which came into 
operation at a period immediately antecedent to the present, and are still 
in progress. I am now obliged to interrupt for a time the historical 
sequence of this account, in order to discuss certain special points, par- 
ticularly the strand-lines of the north and the much-debated problem 
of the temple of Serapis near Pozzuoli. Next, we must consider the 
different behavioiu* of incompletely closed seas, such as the Baltic and 
the Mediterranean, and then we shall be able to return to the most recent 
transgressions. 



326 



CHAPTEB rni 

THE STRAND-LINES OF NORWAY 

From Tjoalma Yagge to the sea. From lake Tonie& to the sea. Movement of the 
ice contraiy to the slope of the valleTs. Origin of the glint lakes of Lapland. Ancient 
strand -lines of the Qords. Origin of the seter. Comparison with Greenland. Vestiges 
left by the retreating ice. 

1. jFVom Tjoalma Vagge to the sea. The coasts of northern Norway 
are marked at intervals by long terraces or grooves cut in the rocL 
The origin of these terraces and grooves is a question that cannot fail to 
suggest itself to any observer visiting this region. Since Leopold von 
Buch's memorable journey at the beginning of the last century, and 
Bravais' observations on the rise of the strand-Hne in the Alten fjord, the 
problem has been made the subject of an extensive literature. The 
strand-lines in their various forms are not the only problem which here 
claim attention ; another and scarcely less disputed question is associated 
with it — ^the origin of the fjords. A glance at the map shows us that un- 
mistakable relations exist between the great lakes of Lapland and some 
of the valleys which open into the fjords towards the west ; some of these 
lakes themselves actually discharge their water into the Atlantic OoeaiL 
These facts show that we cannot hope to reach definite conclusions as to 
the phenomena which occur at the mouth of a fjord without a knowledge 
of its entire valley system from the existing watershed down to the sea. 

I have already mentioned in a previous chapter that in attempting 
a general survey of the connexion between these phenomena I should choose 
as illustrations those valleys possessing outflows which unite in the Maals EIv 
and discharge, above lat. 69° N., by a long estuary into the fjord of Malangen 
(II, p. 58). Leaving the two points we had reached on the watershed, 
I will now attempt to discuss, after a brief glance at the map (PL 1), the 
peculiarities of the valleys and the features which characterize the shored 

The Maals Elv is formed by the waters of three main valleys ; these are 
the Buosta, the Divi, and the Bardo Elv. 

' The heights marked with an asterisk "were determined hy means of an excdlent 
aneroid and compared with the observations of the meteorological station at TromaO. 
The Director, Herr Hann, was good enough to make the calculations. Other data ss (o 
height have been taken from the official map published by the Norwegian Govemmeiii 
In this chapter measurements are frequently given in Norwegian feet ; this was almost 
inevitable on account of the frequent use of whole numbers by the observers (1 Norwegian 
foot B 0-8187 meter; 1 mile « 11,295-8 meters). 



CH. vm] 



VALLEYS IN TROMSO STIFT 



327 



The Ruasta, which lies moet to the east, is discharged from the glint 
lake, Rnosta Jaare ; it flows at first in a gentle curve to the north-west 
and west-north-west» and traverses the Lille Buosta-Vand, at the foot of 
the Buoeta i^eld (II, Fig. 6, p. 60). From this lake onwards the course of 
the Maab Elv, so long as it maintains a westerly direction, should evidently 
be regarded aa the continuation of the Buoeta. Beyond the bend of the 
Maals Elv, the direction of the Buosta is continued in an unmistakable 
manner through the elongated lake of Ands-Vand (153-7 meters) into the 
Solberg fjord. Thus the waters of the Buosta are diverted under existing 
conditions and flow through the lower Maals Elv into the Malangen fjord. 

The middle river, the Divi, is connected by a secondary valley on its 
right side, the Skakterdal, with the high plateau of Tjoalma Vagge, and in 
that direction no watershed appears to exist between the river and the 
glint lake, Tjoalma Jaure. A very remarkable feature is presented on the 
left side of the Divi by two valleys, which descend at first in a directi(m 
opposed to that of the Divi, and conclude by entering it, after swinging 
round in a sharp bend; these are the valleys of the Anavanda and the 
Hogskar EIlv. Near Overgaard, close to the Buosta Vand, the Divi 
reaches the upper Maals Elv, which here, as we have seen, is the continua- 
tion of the Buosta, and its waters turn to the west almost at a right angle. 
But the direction of the Divi to the north-north-west is clearly continued 
in the chain of larger and smaller lakes, which includes the Fjeld Frosk, 
Tagvand (220 meters) and Sagelvand (90 meters), and terminates in the 
Sorkjosen in the Bals fjord and the Nord fjord in Malangen. Thus the 
Divi, like the Buosta, is twice diverted from its original course ; the direction 
of which is continued by a chain of lakes standing at a higher level. 

The third river, the SardOj includes two reaches, an upper and a 
lower. The upper runs to the north-west, and its higher part is the glint 
lake, 48 kilometers long, which arises from the junction of the Lonnes 
Jaure (Lejna Vand) and the Alte Vand (516 meters). This great sheet of 
water, which is evidently only an enlarged river course, flows to the north- 
west towards the Atlantic Ocean, but parallel to it and close beside it lies 
the Gievne Jaure, 16 kilometers long, which flows out towards the south- 
east^ like most of the glint lakes which discharge into the Baltic, but in 
this case the outflow bends round in a sharp curve to join the Lonnes Jaure, 
and thus precisely repeats the behaviour of the lateral valleys on the left 
side of the DivL This upper reach of the Bardo receives the northwards- 
flowing Sordal Elv, which runs in a deep bed from the watershed close to 
the lake of Tomea. Near the parish of Bardo the north-westerly running 
upper branch of the Bardo Elv terminates» and the valley takes a north- 
north-east direction. At this place there is a veiy low col, which passes 
from Bardo westward to the Salangen Elv, the upper lake of Salangen, and 
the Salangen f)ord. But the river itself now enters its lower reach below 



328 THE STRAND-LINES OP NORWAY [pakthi 

Bardo, rans to the north-north-east, and encounters ahuost at right angles 
the direction of the Ruosta ; while the lower Maals Ely, together with its 
estuary, proceeds to the north. 

Thus our first glance at the map reveals a peculiarity of fundamental 
importance in the study of this r^on, where micceagive valley systems haxe 
been superi/mposed one upon another. The river valleys are continued in 
long troughs, which are occupied by lakes and lie several hundred feet below 
them ; thus the Ruosta is continued into the Solbeig fjord, the Divi with 
equal clearness into the chain of lakes which extend from the Tag Vand to 
the BaJs fjord, and the upper Bardo in all probability to the west, towards 
Salangen. I may at once mention that these higher lying basins are all, 
so far as I am acquainted with them, without exception renmants of 
excavated glacier beds. The ice filled all the hollows of the mountains 
and moved forwards in them, Imt the running water which supervenei 
selected orUy a part of the glacial valley system, and deepened that only. 
The remainmg part, lying at a higher level, has been left behind in broken 
chains of lakes. 

We are now able to interpret these valley systems, different in age and 
disposed at different levels. The oldest and least known is the pre-glacial 
valley system. Fragments of it are represented by Anavandene, Hogskar 
Elv, and Gievne Jaure, where the ancient slope to the south-east is pre- 
served. With the advent of the glacial epoch the ancient channels were 
filled with ice, and enlarged after a plan which depended on the directicm 
and damming back of the ice, but not on the fall of the pre-existing valleys. 
Thus the second or glacial valley system arose. The third is the existing 
system, which the rivers are still shaping out. 

Nearly the whole region of these valleys is included within the flat- 
bedded ancient table-land, the structure of which has already been described. 
The band of gabbro and eclogite, which strikes from Lyngen obliquely 
across the valley of the Maals Elv towards the Istind in Bardo, appears 
to have had no influence on the formation of the valleys. The table-land 
terminates on the Swedish side in that steep descent we have termed the 
Lapland glint, and the great lakes are true glint lakes lying across the 
slope; thus the glint runs across the middle of lake Tome&, across 
the Qievne Jaure and the Alte Vand. In the valley of the upper Divi the 
ancient foundation, according to Pettersen, extends somewhat further to 
the north ; on Tjoalma Vagge the glint is broken up into great bastion- 
like table mountains (11, Fig. 7, p. 61), and thence it proceeds further to the 
Buosta Jaure. Thus these lakes, whether they discharge into the Baltic 
sea or the Atlantic Ocean, follow the general rule and extend transversely 
across the glint. 

The movement of the ice, however, so far as I could discover, was 
everywhere directed towards the Atlantic: more to the south it needed 



CH. Ym] ICE POETALS OF TJOALMA VAGGE 329 

the laborious investigations of Horbye and his successors to show that the 
highest level of the ice-mass lay east of the existing watershed ; but here 
in the north the fact is obvious at first sight, and if any doubt were felt 
it would be at once removed by the conspicuous blocks of red Swedish 
granite which have been carried in inunenee numbers acnwa the glint, and 
through its portals, on to the Norwegian table-land. 

Let us now examine the country a little more closely. 

Our first traverse runs from Ijoalma Vagge through the Skakterdal 
and Divi across the chain of lakes of the Tag Vand and Sagel Vand to the 
BaLs Qord. This is the same line as that we have already described, at 
least as regards its more elevated part in a previous chapter (II, p. 60). 

On the lofty barren surface of the Ijoalma Vagge we are surrounded 
by a boundless labyrinth of low mounds of stones. Peat bogs and number- 
less pools of water broaden out between them. Low rounded bosses of 
red granite project from the mass of d^ris, and reveal the fact that we 
are standing on the foundation of the table-land. This is moraine land. 
No one can say exactly where the watershed lies between the Baltic sea 
and the Atlantic Ocean, the latter as the crow flies only 77 to 80 kilometers 
away. In the season when the snow melts all this country is probably 
inundated ; nothing but a few stony ridges, higher than the rest, will rise 
above the water here and there, and the direction of the outflow will 
depend upon the wind. 

Towards the north, west, and south-west rise like phantoms the great 
cubical masses into which the glint is broken up ; far away the outlines 
of the table mountains are veiled in mist, and through the mist their 
snowfields glisten. The foot of the nearest mountain, Store Jerta, is 
surrounded by a terrace, about 15 to 20 meters high. Whether this little 
diflT, which faces the labyrinth of stony mounds, indicates the border of 
a once continuous and permanent sheet of water or of a temporary 
inundation is a question I will not pretend to answer. (Upper edge, 
728-4 meters.*) 

A chain of pools, united by a thread of water, extends towards the 
portal between the table mountains Store Jerta and Namna. The breadth 
of this is scarcely more than 800 meters, at the level of the route by which 
we pass through it, and this is no great height above the bottom. We 
stand here at one of the passages through which the ice advanced towards 
Norway. To-day it presents the aspect of a glacier bed, scarcely changed. 
Eemnants of the moraines of Ijoalma Vagge appear to penetrate into the 
portal ; they are broken up into barren isolated sand-hills with somewhat 
steep slopes scattered here and there over the ground. Higher up on the 
slopes of both mountains, particularly of the Namna, we see several long 
mounds of sand, more like dunes than moraines, which evidently slope 
downwards, that is, with the water drainage. At one spot we see five or 



330 THE STRAND-LINES OP NORWAY [pabtih 

six of these ridges on the Namna, resting obliquely one against the other. 
They are the vestiges of the last movement of the ioe in the pass. Herr 
Pettersen has ascended the Store Jerta and found blocks of the red granite 
from Sweden on its summit ; there was thus a time when the great pylons 
of the portal were overwhelmed by the ice. 

Where the land widens out below and at the mouth of the pass, 
hundreds of circles, two to four meters in diameter, are scattered over the 
surface ; this is enchanted ground, a dancing place for the elves ; so sack 
spots figure in folklore : but in fact the circles are formed by ccdonies of 
grass, which grow outwards at the margin and die away in the middle, 
a sort of vegetable atoll ^. 

We have now walked round the Store Jerta ; the finer d^lnris, and with 
it the fairy circles, is gone, and we stand on a broad glacier bed. In 
a little hollow are seen the first stunted birches (699-6 meters^) ; beyondSt 
polished surfaces extend far and wide. These are great moutmmM 
bosses of schist and quartzite, and although the whole region is flat- 
bedded, yet violent local foldings are revealed on the polished sur&oes of 
the rocka Over the roches moutonn^ erratic blocks are scattered in a 
strangely uniform manner, we might almost say at equal distances. The 
blocks consist chiefly of red granite : as a rule they are somewhat less than 
a cubic meter in size ; some of them larger. They lie on and between the 
roches moutonn^ just as the ice has left them, and sometimes the presBOie 
of the hand is sufficient to send a heavy mass clattering do¥naL from the 
summit of the boss on which it rests. The moutonn^ed surface maybe 
recognized on the slope at our right for a height of at least 100 mete» 
above us. 

On this part of the glacier floor sand and small stones are rare, and the 
regular distribution of the blocks on the roches moutonn^es does not 
correspond with the ideas generally entertained as to the nature of a 
ground moraine. While here everything looks so fresh and undisturbed, 
on our right the brook which issues from the portal, and in the interval 
has become the raging Skakter, has already excavated in the same rocky 
ground a ravine more than 100 feet deep. 

We now proceed along the gentle slope of the glacier bed, when it 
suddenly comes to an end. A low cliff of flat-bedded quartrite follows; 
below this a spring, and then a steep wooded descent into the valley of the 
DivL At the bottom we see the cone built up of the material which the 
Skakter has brought out of the ravine, and then the rushing DivL 

^ Profesaor Ton Eemer tells me that Sesaleria eoerulea is the plant which geneially 
forms these fairy rings, and that the destruction of the vegetation within the green 
border is produced in two ways : in the case pf plants with radially extending ifaiiomes 
the older parts of the rhizome die and decay at the centre, and new plants do not imme- 
diately take their place ; secondly, the meadow grasses are killed by the mycelinm of 
a fungus which accompanies the radial growth. 



CH* ^VUl] 



THE DIVI ELV 



331 



The Dividal is a rectilinear narrow valley ; its whole length amounts 

to about 60 kilometers, of which 86 go to the reach below the mouth of 

the Skakter valley. The hut of Frihedsli (187 meters*), stationed in the 

virgin forest of the valley bottom, was the starting-point of our wanderings. 

Here in August we found flowery meadows, a noble forest growth of birch, 

botterflies, and in the short twilight even bats, and only the trifling height 

to ^which the wood extends up the mountains reminded us that we were in 

the latitude of central Greenland. The mountains which surround the 

middle Dividal are 4,000 to 6,000 feet in height (Njunnes Vane opposite 

Frihedsli, 1668-6 meters), and throughout the year are covered with more 

or less extensive sheets of snow. If we ascend the slope of the Anaskole 

above Frihedsli we find that both sides of the valley up to a certain fairly 

uniform level are formed of a continuous sheet of d^ris : I put its height 

above the valley bottom at 200 to 250 meters. In this d^ris we also find 

fragments of red Swedi^ granite : we might regard it as a marginal 

mcnraine, but it nowhere rises in mounds from the rocky walls of the 

valley. 

Through the narrow valley bottom rushes the Divi El v. I do not think 
it would be exaggerating to say that the volume of its waters is probably 
six to eight times as great as would be met with under similar conditions 
of topography in a vaUey of the Alps. The high rainfall, the great 
quantity of snow which accumulates during the long winter, and the 
length of the day in summer are responsible for this extraordinary volume 
of water, and explain the great and unnfloal power of eioeion. 

After a walk of some hours down from Frihedsli we reach a somewhat 
broader vaUey bottom, and here the first settlements have been made. The 
river is bordered by r^ular terraces ; two or even three occur one above 
the other. They have undoubtedly been formed by the river itself ; this 
is apparent from the fact that ox-bow terraces, that is islands of alluvium 
which have been left in place, occur on the older terraces. Such isolated 
spurnshaped fragments of terraces can only have been left by running 
water, and indeed by the displacement of the river bed. They overlook 
the adjacent valley bottom ; several of the farms have been built on the 
terminal spur of one of these ox-bow terraces. 

Towards the end of the valley the Divi enters a ravine ; on the left we 
ascend a ridge, presenting in its lower part moutonn^ed bosses of schist 
and quartzite ; above, an accumulation of blocks, probably part of a moraine, 
which, however, lies higher than the valley bottom of tiie Divi ; on the 
other side the road runs down towards Overgaard (58*6 meters^), where 
the Divi joins the Maals Elv; here also river terraces occur. 

We cross the low-lying terraced river valley and enter the ancient 
prolongation of the Dividal toward Baia fjord indicated by the lakes. 
This, as we have seen, is a glacier bed. Pettersen has also recognized the 



332 THE STRAND-LINES OF NORWAY [pamih 

fact that the ice which came down the Dividal crossed the valley of the 
Maals Elv and proceeded towards Bals fjord, and consequently that 
the valley of the Maals Elv has been deepened subsequently by erofiion ; 
he showed besides that on the right side of this glacier bed blocks of red 
granite occur nearly up to the summit of the Omasvane, and on the left 
they cover the slopes of the great Mauktind, which is much higher than 
the Omasvane, to about 2,500 feet above the sea, or as far as it has been 
examined K The ice as it passed through this trough must thus have 
been of great thickness. The terraces of the .river valley have now dis- 
appeared ; roches moutonn^ surround us on all sides. Their smooth and 
rounded summits protrude from the peat bogs which surround the green 
waters of the Lompol Vand, and accompany us as far as the islet-strewn 
waters of the Tag Vand (178-8 meters), which wash against the foot of the 
Mauktind ; the rocky slopes, as we can plainly see, are likewise polished to 
a great height above us. Beyond the Tag Vand the country presents 
peculiarly characteristic features. On the slopes of the Omasvane we 
wade through high ferns ; a few beeches remind us of our own land : but 
on the left lies the bare U-shaped, polished glacier bed with its roxmded 
bosses, and in between them innumerable little pools of water which drain 
partly into the Tag Vand and partly into the Sagel Vand to the north ; 
they are the most remote retreats to which the salmon ascend. No channel 
of any depth b to be seen; travelled blocks are strewn around: it is 
absolutely impossible to suppose that this glacier bed can have been 
covered by the sea -since the retreat of the ice, and that the scene should 
have remained so completely unchanged, the glacial striae so fresh, the 
little hollows between the bosses unsUted up, and the scattered blocks 
undisturbed, notwithstanding the play of the tides. And yet this glacier 
floor here lies only 130 to 150 meters* above the sea, and descends lower 
and lower as we proceed. 

Above the farm of Myre, nearly 400 feet above the sea, we still find 
the hard schists quite sharply marked with glacial striae directed down the 
valley, and below the farms the roches moutonn^ and travelled blocks, 
amongst them red granite, lift their heads above the peat bog, which 
stretches away as far as the Sagel Vand. The bed of the glacier has 
sloped gradually down to the level of the Sagel Vand (90 meters). 

The Sagel Vand is the last lake that we meet with along this traverse ; 
it is said to be 220 feet in depth, so that its bottom lies a little below the 
existing level of the sea. On its right we observe what appears to be a bit of 
lateral moraine, and at its lower end are indications of a terminal monune, 
but as to this I am not certain. With this lake the U-shaped valley comes 
to an end. Recent alluvial land evidently forms a considerable part of its 

^ E. Pettenen, Det nordlige Norge under den glaciale og post glaciale Tid: IHt 
Granitisk flytblokkestrOm ndefter Balsiljorden ; TromsO Mas. Aanh., 1884, VII, pp. 1-1^ 



CH. vni] FROM LAKE TORNEA TO THE SEA 



333 



lower termination, although rounded bosses of rock project from it. The 
allnvial land, formed of sand and gravel, extends below the lake in a plain, 
which broadens out downwards : we proceed along this plain for a distance 
of 4-5 kilometers, and then before us lies the Bals fjord. We now find oiur- 
aelves on the upper edge of a slope, which towards the sea is divided into 
tw^o sharply defined terraces. 

A result of great importance follows from all these observation& The 

ice-streams which entered the country from Sweden, flowing through the 

portals of the glint and above them, had their bed at a higher level than 

a great part of the existing Dividal and the existing valley of the Maals 

Elv. We have traced this bed through the upper SkakterdfJ, as far as the 

steep descent to the Dividal. Its continuation ia seen in the U-shaped 

trough of the Tag Vand, where it is much deeper than in the portal, but 

still some hundreds of feet above the existing valley. The slope of this 

glacier is directed towards the Bals fjord. Its pdiahed bed, after He level 

has fallen to 90 meters, sinks beneath the waters of the Sagd Vand; ihis is 

held up on the seaward side by a dam, formed perhaps by a terminal 

TnoraiTte,